Search

SAMMENDRAG Introduksjon. Komfort og stabilitet er viktige faktorer ved tilvirkning av protesehylse til amputasjonsstumpen hos personer med amputasjon av underekstremiteten og påvirkes av trykk, forskyvning og friksjon. Flere suspensjonssystemer finnes, men en systematisk objektiv vurdering av suspensjonssystemene med hensyn til rotasjonsmotstand mangler. Målet med studien var å finne forskjellen i rotasjonsmotstand mellom protesehylse og amputasjonsstump for fem ulike suspensjonssystemer og fire ulike aksiallaster. Metode. Testriggen besto av en gjennomsiktig hylse, en kunstig kopi av amputasjonsstumpen i silikon (mock) med en tekstilliner utenpå. Et kamera festet på hylsen målte rotasjonsmomentet (slippmomentet) ved 1 mm forskyvning av liner relativt til protesehylsen i en rotasjonsbevegelse. Aksiallast og trykk ble også målt. Ti undersøkelser med fem suspensjonssystemer ved fire aksiallaster (20, 40, 60 og 80 kg) ble utført. Alle tester var gjennomført med låsepinne festet til lineren. De fem suspensjonssystemene som ble testet var «ikke-vakuum», «sleeve-passivt vakuum», «sleeve-aktivt vakuum», «elastomerdekket liner-passivt vakuum» og «elastomerdekket liner-aktivt vakuum». Resultater. Den høyeste gjennomsnittlige verdien på slippmoment var 6,1 Nm og ble funnet med «sleeve-aktivt vakuum» ved en aksialbelastning på 80 kg. Det laveste slippmomentet var på 0,6 Nm med «ikke-vakuum» ved 20 kg. En ANOVA-modell med slippmoment som avhengig variabel og suspensjonssystemer og aksiallast som faktorer viste signifikante forskjeller på suspensjonssystemer med økning av aksiallast, og en forskjell mellom faktorene. Post-hoc-tester viste at «sleeve-aktivt vakuum» og «elastomerdekket liner-passivt vakuum» hadde høyest rotasjonsmotstand, etterfulgt av «elastomerdekket liner-aktivt vakuum», «sleeve-passivt vakuum» og «ikke-vakuum». Alle forskjellene var statistisk signifikante (p

National Joint PhD Programme in Nautical Operations and the Department of Ocean Operations and Civil Engineering Remote operations have grown dramatically in popularity in recent years across countless industries thanks to technological advancements that enable human operators to monitor and control machines from distant locations. Remote operations are widely considered to be a tool that improves the performance and efficiency of ship operations. However, simply relocating humans from onboard a ship to a shore control center (SCC) and replacing existing ships with autonomous or unmanned ships does not prevent accidents or improve the efficiency of ship operations. Humans significantly affect performance, even with remote operations. This reality underscores the need for companies to employ new approaches to assessing human factors within remote ship operations. This thesis studies the human factors behind remote operations with a focus on those that impact performance and efficiency. This thesis proposes the novel smart training in remote operations (STiRO) methodology to predict the status of human factors, including mental workload and stress during remote operations, using machine learning (ML). The proposed methodology produces an interpretable ML model that accurately correlates the brain activity of human operators with the level of mental workload and stress that they perceive during remote ship operations. In this methodology, an electroencephalogram (EEG) device records the brain activity of human operators, which the ML model uses to predict human operators’ levels of stress and mental workload. In fact, interpretable ML models allow observers to infer the human factors that contribute to mental workload and stress levels. In this way, training programs can utilize the STiRO methodology to identify human factors that influence operator performance and provide customized training programs based on individual differences. In effect, the STiRO methodology enables remote operations’ stakeholders to replace traditional technical training with efficient, human-centered training and, in turn, to mitigate human factor issues during remote operations. In addition, SCC designers can utilize the STiRO methodology to identify human-human and human-machine interactions that transform into human factors issues while standardization organizations regulate SCC designs to mitigate human errors.

Flexible graphite (FG) is obtained by compaction of exfoliated graphite in the form of sheets having a thickness of 0.2 to 3 mm and density ρ from 0.7 to 1.9g/cm3. Sigraflex® is a commercial type of FG with ρ = 1 g/cm3 employed as beam dumping material in the Target Dump External (TDE) cores of the Large Hadron Collider (LHC) located in Geneva, Switzerland. The performances of the TDE components are assessed by comprehensive FE simulations whose accuracy depends on the material parameters assigned to each component. In such context, this thesis aims to find the model and the necessary parameters for the mechanical simulation of the Sigraflex® core in the TDE. At first, a review of Sigraflex® and FG properties available in literature will be presented: this will include the description of the microstructure, the collection of the mechanical and thermal properties, and their critical comparison with other well-known types of graphite. Then, an extensive experimental campaign will be reported. For the first time, the application of a Focused Ion Beam - Scanning Electron Microscopy (FIB-SEM) technique to FG for the quantitative investigation of the pores’ sizes and shapes and the micro-sheets’ thickness and arrangement, will be described. Moreover, the experimental setup and the mechanical properties obtained by in-plane tension, out-of-plane compression and nanoindentation tests will be outlined and thoroughly discussed. In particular, nanoindentation revealed to be an easy-to-use method to measure the orthotropic elastic properties of FG. Finally, although the focus will be mainly on the mechanical properties, a thermomechanical testing campaign will be presented, too. Based on experimental observations, the behavior of FG will be assimilated to other well-known materials such as crushable foams and crumpled materials. A 1D analytical model will be proposed to decouple the deformation contributions from the graphite-like and crumpled-like behaviors, and the extension of the nonlinear stress-strain behavior to 3D cases will be discussed. A practical solution will be to use a material model found in Ansys LS-Dyna library i.e., MAT_142 transversely crushable foam. Its suitability was investigated by means of 3D FE simulations of nanoindentation and the numerical force-displacement curves will be shown in comparison with the experimental curves. Although not exhaustive, the results are promising and can be considered as a reference point for future works on FG characterization.

Prototyping has become widely adopted across engineering design practices and is suggested practical and reflective advantages in new product development. Although this has yielded increased research focus, holistic and empirically grounded insight into —what prototyping entails in projects, how prototyping is facilitated, and why prototyping should be encouraged is nevertheless limited. This thesis involves an industrial research project joint between Laerdal Medical and TrollLABS NTNU, regarding prototype-driven development in the fuzzy front end. The project considers the context of simulation-based medical training equipment and investigates how prototyping contributes to their development. This thesis includes 11 academic contributions and 10 case projects and initiatives presented and utilized in a binding article. The research project has pursued and addressed three goals: 1.) Explore new product opportunities within the healthcare training domain. 2.) Develop and utilize new technology for healthcare training applications. 3.) Enhance process knowledge and methodological foundation on which Laerdal Medical, as the industry partner and stakeholder, develop new concepts and technology. The collective research output of case projects, activities, and academic contributions address the two first goals, while the binding article of this thesis addresses the third. Chapters 4, 5, and 6, each viewed from an organizational level, —describe prototyping to develop new medical training equipment, facilitate prototype-driven development by wayfaring, and expand organizational knowledge by prototyping. The chapters address what, how, and why we prototype in industrial and research settings, and they account for three individual learning loops. Respectively, the chapters present empirical insights for operational, tactical, and strategic use of prototyping in the fuzzy front end of new product development. Sammendrag Prototyping har blitt utbredt I design- og produktutviklings-praksis, og er sett på som et nyttig verktøy for å skape og lære under utviklingen av nye produkter. Selv om det har blitt forsket på prototyping, mangler man en helhetlig forståelse av hvordan prototyping brukes I prosjekter, hvordan man kan legge til rette for prototyping, og hvorfor prototyping bør fokuseres på. Denne oppgaven beskriver et samarbeidsprosjekt mellom Laerdal Medical og TrollLABS NTNU, som omhandler prototypedrevet utviklingen i Fuzzy Front End. Prosjektet har involvert utviklingen av medisinsk simulering- og trenings-utstyr, og hvordan prototyping bidrar til denne utviklingen. Oppgaven inneholder 11 vitenskapelige artikler og 10 case -prosjekter og -initiativer som blir brukt I en sammenfattende avhandling. Forskningsprosjektet har følgende mål: 1.) Utforske nye produktmuligheter innen medisinsk opplæring og trening. 2.) Utvikle og nyttiggjøre seg av ny teknologi og applikasjoner for medisinsk opplæring og trening. 3.) Styrke kunnskapsfundamentet som samarbeidspartner og prosjekteier, Laerdal Medical, bruker til å utvikle nye konsepter og teknologi. De samlede forskingsresultatene, i form av case-prosjekter, aktiviteter og akademiske bidrag, tar for seg de to første målene, mens denne avhandlingen vil ta for seg det tredje. Kapittel 4, 5 og 6, alle sett fra ett organisasjonsnivå, beskriver prototyping for å utvikle nytt medisinsk treningsutstyr, tilrettelegge for prototypedrevet utvikling ved hjelp av wayfaring, og forbedre organisasjonskunnskap og læring ved prototyping. Kapitlene tar for seg hva, hvordan og hvorfor vi prototyper i industrielle- og forsknings -miljøer, og de står for tre individuelle læringssløyfer. Henholdsvis presenterer kapitlene empirisk innsikt for operasjonell, taktisk og strategisk bruk av prototyping i Fuzzy Front End av produktutvikling.

Verdens helseorganisasjon anslår at omtrent 685 000 kvinner over hele verden døde av brystkreft i 2020, hvorav en uforholdsmessig del var kvinner i utviklingsland. En stor del av disse dødsfallene kan forebygges gitt økt bevissthet og tidlig diagnose. Denne masteroppgaven presenterer utviklingsprosessen for en prototype av en palpasjonstrener basert på det nye haptiske grensesnittkonseptet av ferrogranulære partikler. Denne prosessen inkluder utforskning, prototyping og testing. Palpasjon er den medisinske undersøkelsesprosessen som involverer medisinsk personell som bruker hendene og følesansen for å undersøke den underliggende kroppsdelen. De ser etter uregelmessigheter og taktile forskjeller som endringer i hardhet, tekstur og smerterespons. Treningsapparatene som brukes i det medisinske feltet i dag tilbyr ikke fleksibiliteten og justerbarheten som kreves for å levere et varig og overførbart ferdighetssett. Derfor utviklet jeg en adaptiv palpasjonstrener basert på konseptet granulære partikler, brakt fra feltet myk robotikk til feltet medisinsk simulering. En justerbar pumpe brukes til å trekke et vakuum i et forseglet kammer fylt med det magnetiske granulatet. Granulatet kan plasseres fritt ved hjelp av en magnet, og brukeren har full kontroll over formene som lages. Den resulterende overflaten kan ha forskjellige tekstur- og hardhetsegenskaper. Det gjennomføres et eksperiment med 6 deltakere for å teste konseptets evner. Eksperimentet vurderer deltakernes evne til å korrekt lokalisere og beskrive de konstruerte uregelmessighetene. Resultatene fra eksperimentet indikerer prototypens evne til å gjenskape former med høy presisjon, simulere dybde og lære nybegynnere palpasjonsferdigheter. The World Health Organisation estimates that approximately 685 000 women worldwide died from breast cancer in 2020, with a disproportional part of which were women in developing countries. A great number of these deaths are preventable given raised awareness and early-stage diagnosis. This master's thesis presents the development process of a palpation trainer prototype based on the novel haptic interface concept of ferrogranular jamming, including research, exploration, prototyping, and testing. Palpation is the medical examination process involving the medical professional using their hands and sense of touch to examine the underlying body part. They look for irregularities and clues such as changes in hardness, texture, and pain points. The training devices used in the medical field today do not offer the flexibility and adjustability required to deliver a lasting and transferable skill set. Therefore, I developed an adaptive palpation trainer based on the concept of granular jamming, brought from the field of soft robotics to the field of medical simulation. An adjustable pump is used to draw a vacuum in a sealed chamber filled with the magnetic granulate. The granulate can be freely positioned using a magnet, and the user is in total control of the shapes created. The resulting surface can have different texture and hardness characteristics. An experiment is conducted with 6 participants to test the capabilities of the concept. The experiment assesses the participants' abilities to correctly locate and describe the irregularities constructed. The results from the experiment indicate the prototype's ability to recreate high-fidelity shapes, simulate depth, and teach novices the skills of palpation.

Denne masteroppgaven beskriver utviklingen av et par med smartsko for å måle ballistokardiografi. Masterprosjektet er en fortsettelse av en prosjektoppgave som lagde en prototype som et konseptbevis. Målet med masteroppgaven var å lage en smartsko som måler ballistokardiografi, samt validere resultatene gjennom testing. I tillegg, skulle alle «serendipitous» oppdagelser bli utforsket. Skoene ble utviklet gjennom en iterativ metode med utgangspunkt i Wayfaring med prototyping som hovedverktøy. Prototyper ble laget for teste og validere resultater, samt fungere som et verktøy for å generere kunnskap og design spørsmål. Både divergerende og konvergerende prototyper ble laget for å nå den beste løsningen. I tillegg, ble en småskala testoppsett laget som simulerte ballistokariografi for å kunne raskt teste og validere påvirkningen til diverse design parametere. Det endelige konseptet, Sole 2.0, består av et par med joggesko med fem væskefylte blærer i hver sko. Hver av blærene er koblet til en trykksensor. Smartskoen måler trykket under foten til brukeren og fanger opp bevegelsen til kroppen som følge av blodsirkulasjonen. I tillegg, er Sole 2.0 designet for å kunne kobles til en ballistokardiografi vekt som fungerer som en referanse og en PPG-sensor for å måle pulstransporttid og muliggjøre segmentering. For å validere resultatene ble et forsøk gjort med 14 deltakere. Deltakerne ble i målt tre perioder med en hvileperiode mellom test to og test tre. I testperiode to ble en «coldpressor» test utført for å øke blodtrykket til deltagerne. Blodtrykket ble økt for å gjøre det mulig å beregne pulstransporttid. Resultatene fra forsøket er ikke behandlet, men foreløpige resultater er svært lovende. To ut av 14 datasett er behandlet og ballistokardiografi er suksessfullt målt i alle testperiodene for de to datasettene. I tillegg er den beregnede pulstransporttiden korrekt invers relatert til blodtrykket. This master thesis describes the development process of a pair of smart shoes for measuring ballistocardiography. The thesis was a continuation of a proof-of-concept prototype from a previous project thesis. The thesis aimed to develop a functioning smart shoe for measuring ballistocardiography and validate the results through testing. In addition, any serendipitous findings in the process were to be explored. The shoes were developed using an iterative approach based on wayfaring with prototyping as one of the main tools. Multiple prototypes were created to validate hypotheses and act as a tool for generating new design questions. Both divergent and convergent prototypes were created to ensure the best solution was created. In addition, a benchtop setup was created for simulating the ballistocardiogram to enable rapid testing of the impact of changing design parameters. The final proposed concept, called Sole 2.0, consists of a pair of running shoes with five water-filled bladders in each shoe, each connected to a pressure sensor. The smart shoe measures the pressure under the user's feet and captures the body's movement due to the ejection of blood. In addition, the sole 2.0 allowed a BCG scale to be connected to act as a reference for the captured BCG and to attach a PPG sensor for calculating pulse transit time and aiding in segmentation. To validate the concept, a test was run with 14 participants. Participants were measured in three test periods with one resting period between tests two and three. A cold pressor test was used to increase the participants’ blood pressure during test period two. Blood pressure was increased to enable the calculation of pulse travel time. The results from the testing are not fully processed. However, the preliminary results from two of the fourteen datasets show great promise, with the BCG being captured in all measurement periods and the inverse relationship between pulse transit time and blood pressure being shown.

Temperaturstyring i Smeltet FilamentStyring (SMS) blir mer relevant for forutsigbar styrke i Additiv Tilvirkete (AT) komponenter. Ved å overvåke lagtemperaturer under prosessen av SMS, kan styrke bestemmes ved å sammenligne lagtemperatur med styrke for en gitt konfigurasjon. Metoder for å bestemme lagtemperaturer i prosessen er fortsatt under utvikling, og berøringsfrie infrarøde temperatursensorer blir mer anvendelige. Med økende følsomhet, raskere behandlingstider og mer pålitelige temperaturkalibreringer, viser berøringsfrie infrarøde temperatursensorer lovende funksjoner for lagtemperaturm˚alinger i SMS. Ved å bruke en svært nøyaktig berøringsfri infrarød temperatursensor ble dette prosjektet i stand til å overvåke lagtemperaturer i prosessen i detalj. Hvert lagmønster ble lett identifisert, og hver bevegelse i fabrikasjonen kunne observeres. Resultatet ble en lagtemperatur modell av prøven som ble brukt til forsøkene. Potensialet til fulltemperaturmodeller er introdusert, men med en usikkerhet om emissivitetseffekt. Gjennom eksperimenter med den kontaktlse infrarøde temperatursensoren og termoelementer i kontakt med materiale kunne endringer i emissivitet forutses. Dette prosjektet introduserer bruken av en presis berøringsfri infrarød temperatursensor for å overvåke lagtemperaturer i prosessen av SMS og effekten av emissivitet for dynamiske temperaturer i SMS. emperature management in Fused Filament Fabrication (FFF) is getting more relevant considering predictable strength in Additive Manufacturing (AM) components. By monitoring layer temperatures in FFF in-process, strength can be determined by comparing layer temperature to strength for a given configuration. Methods to determine layer temperatures in-process is still under development, and non-contact infrared temperature sensors are becoming more applicable. With increasing sensitivity, faster processing times, and more reliable temperature calibrations, non-contact infrared temperature sensors show promising features for layer temperature measurements in FFF. Using a highly accurate non-contact infrared temperature sensor, this project was able to monitor layer temperatures in-process in great detail. Each layer pattern was easily identified, and each movement in the fabrication could be observed. The result was a layer temperature model of the specimen used for the experiments. The potential of full temperature models is introduced, but with an uncertainty of emissivity effect. Through experiments with the non-contact infrared temperature sensor and in-contact thermocouples change in emissivity could be predicted for the material used. This project introduces the use of a precise non-contact infrared temperature sensor for monitoring layer temperatures in-process and the effect of emissivity for dynamic temperatures in FFF.

Denne masteroppgaven er en del av et større forskningsprosjekt i samarbeid med NTNU, Fosenregionen, Maritime Robotics og SFI Autoship. Målet med forskningen er å binde mindre urbane regioner av Fosen med autonome cargo fartøyer. For at ubemannede marine cargo systemer kan nå det kommersielle markedet, er det mange mulige utfordringer som må identifiseres og løses. Autonome fartøyer kan enten fortøye til flytende brygger eller til stasjonære kaier. Vi mennesker har evnen til å enkelt kunne justere våre bevegelser til omgivelsene, mens autonome fartøyer har i dag ikke den samme evnen. I tillegg har mennesker evnen til å lokalisere og posisjonere et fartøy til samme lokasjon og utføre vanskelige løfteoperasjoner mellom fartøyet og kaien. For at autonome fartøyer som skal kunne gjøre den samme prosessen uten noen menneskelig medvirkning, kreves det ulike løsninger som sammen gir autonome fartøyer en lignede evne. Denne oppgaven beskriver en kamera-basert metode som identifiserer objekt-markører på en kai og kalkulerer fartøyets orientering relativt til disse når fartøyet nærmer seg. Markørene gir fartøyet informasjon om dens orientering i seks frihetsgrader relativt til kaien i et lokalt referanse system. Objekt-markørene er uavhengige om det er en flytende brygge eller en stasjonær kai. Metoden gjør at fartøyet kan estimere en lokal presis posisjon av kaien uten bruk av ekstern kommunikasjon til fartøyet. Ettersom kamerasystemet har evnen til å gi informasjon over seks frihetsgrader, kan et slikt system gi autonome fartøyer evner til mer komplekse interaksjoner med en docking stasjon, f.eks. robotarmer på fartøyet som flytter cargo mellom fartøy og en docking stasjon. For å verifisere posisjonsestimatene fra kamerasystemet ble det brukt realtids kinematiske måling med en lokal basestasjon. Denne metoden ga en realistisk og et presist sammenligningsgrunnlag til posisjonsestimatene fra kamerasystemet mot fartøyets globale posisjoner. Posisjonene fra kamera systemet ble i tillegg Kalman filtert i et forsøk på å oppnå bedre nøyaktighet og ytelse, men også integrert som et steg nærmere for å bli en del av navigasjonssystemet til et autonomt fartøy. Resultatene fra de gjennomførte eksperimentene viste at systemet under gode forhold fungerte godt, men i mer utfordrende situasjoner, f.eks. mørke eller reflektert lys, gikk ytelsen til systemet ned. Til slutt, blir det gitt forslag til forbedringer. The focus on autonomy in the maritime industry has in recent years increased. Autonomous vessels have the potential to reduce costs and improve safety in the maritime industry. Unmanned Surface Vehicles (USVs) are reaching the commercial market, e.g., Maritime Robotics, SailDrone, and Kongsberg. However, further enchantment of the field requires more product development to make it possible for USVs to perform more complex operations without any human intervention. This project is a collaboration between NTNU, Fosenregionen, Maritime Robotics, and SFI Autoship. The project's grand vision is to connect urban and remote parts of the Fosen region with autonomous cargo boats. Here, the crew and port handling is considered to constitute a significant cost associated with short-range shipping. The objective of this master thesis is to develop solutions for autonomous cargo handling to connect small remote places, e.g., islands and coastal towns. This master thesis combines theory and practice to verify any predictions. By looking at the theoretical aspects of autonomous cargo handling, in addition to testing, the understanding of the theoretical methods, strengths, and shortcomings is in this project reviewed. This thesis describes a vision-based approach of using fiducial markers to aid USVs while auto-docking. Recent applications of USVs have shown increased ease and increased efficiency of cargo shipping at precise locations. However, further field enhancement requires operations over long periods, e.g., days or weeks of which the vessel must be precisely docked to be refueled/recharged or for complex cargo handling in transit. Utilizing vision-based techniques allows USVs to orient themselves to their environment from their perspective and may represent an efficient method for such vessels to accurately orientate themselves in the docking scenario. In addition, fiducial markers can give USVs six degrees of freedom orientation relative to a target, allowing more complex interactions with a potential pier or a floating docking station. We used RTK GNSS with a base station on the pier to validate the estimated camera positions to obtain reliable additional positioning. The experiments show that essential computer vision (CV) techniques using fiducial markers perform well in obtaining accurate outdoor position estimates in optimal conditions. However, in more adverse conditions, the results demonstrated that the performance decreased significantly. Finally, suggestions for further development are given.

One of the most applied optimization methods in mechanical engineering is topology optimization (TO). The benefits of its integration in the product development process are several, such as reducing material usage in manufacturing, shortening the design cycle, and enhancing product quality. However, the implementation of TO is characterized by the following bottlenecks: the geometrical complexity of its optimized designs, the long optimization times, the sensitivity of its results to the given parameters, and the need for numerous inputs during its workflow. All these issues make TO a complex and time-demanding procedure dependent on designers’ starting guesses and choices during its implementation. It is clear that there is a need for a more automatic and effective optimization procedure. In this thesis, the author uses TO for the weight reduction of structures in mechanical engineering. First, he explores the workflow of the TO and identifies interesting practical aspects in its implementation. The most popular TO-methods, such as SIMP and BESO, are described, categorized, and compared. In addition, the three following TO-practices are developed with respect to the size of the available design space for optimization: TO with limited design space, TO with maximum possible design space, and combined size/shape/topology optimization. The author states that the designer’s choices (inputs) affect the TO-results and categorizes them into five clusters of parameters: design constraints, supports and connections, loads, geometric restrictions due to manufacturing constraints, and software constraints. The sensitivity of the TO-results to the variations of these parameters is explored. Furthermore, different multi-objective, multi-level, and multi-scale optimization workflows are used in the pursuit of the lightest design solutions. To identify the software constraints, a literature review is conducted among the most applied TO-software platforms. As a result, an online library of 70 commercial and open source TO software is developed in the form of a table. This table encompasses the name, company, optimization types, and methods that software uses, as well as its available objective functions and constraints in TO. Moreover, relative research works and representative literature for each software are included. Different 3D models are designed, optimized, and used as case studies to support the theory and tie the academic text to real-world applications of TO. Finally, the educational perspective of TO is checked. The author developed an educational framework of a topology optimization-based learning (TOBL) combining the CDIO-approach and TO. The implementation of the developed TOBL-framework in any study program in CAD-engineering can educate modern CAD-designers to conceive, design, implement, and operate optimized products. The current research work is addressed to practitioners, researchers, teachers, and other engineers looking for new lightweight design concepts. Hence, the aim of this thesis is to provide them, through valuable insights, with a better understanding of TO, as well as to advise them with guidelines and recommendations to avoid common pitfalls.

Norsk sammendrag Denne oppgaven er addresert til forskere og produktutviklere. Den har som hensikt å gi innsikt i løsing av komplekse sensorproblemer i the fuzzy front end av produktutvikling. Sensorer bygger bro mellom fysiske fenomener og prosessering i cyberfysiske systemer og involverer tverrfaglig praksis ved implementering. Som sådan er innovasjonsprosjekter som utforsker bruken av sensorer i the fuzzy front end ofte komplekse. Forskere og produktutviklere må beslutte om problemet krever implementering av sensorer, hva som må måles, hvordan det skal måles, og hvordan de skal teste de potensielle sensorkonseptene. Basert på ti prosjekter og ti tilhørende vitenskapelige publikasjoner, presenterer denne oppgaven innsikt fra observasjoner innen relevante problemstillinger fra sensorteknologi og praksisser innen prototyping og protype testing: Egenskapene og utfordringene til komplekse sensorproblemer i the fuzzy front end utforskes, mulighets skapende prototypeatferd i løsingen av komplekse sensorproblemer, samt verktøy som muliggjør utforskende prototyping av komplekse sensorproblemer i the fuzzy front end. Basert på innsiktene, presenterers oppgavens resultater som åtte hypoteser som kan tjene som utgangspunkt for videre forskning. English abstract This thesis is addressed to researchers and product developers and provides insights on the solving of complex sensor problems in the fuzzy front end of product development. Sensors bridge physical phenomena and computational processing in cyber-physical systems and entail multidisciplinary practices in their application. As such, projects seeking novel and disruptive innovations in the fuzzy front end through sensors are often complex. Researchers and developers need to gauge whether the problem calls for the implementation of sensors, what to sense, how to sense it, and how to test the proposed sensor concepts. Based on ten projects and ten accompanying scientific publications, this thesis presents insights from practical examples for relevant topics within sensor technologies, prototyping practices, and prototype testing practices. Insights are generated on the characteristics and challenges of complex sensor problems in the fuzzy front end, enabling prototyping behaviors for solving complex sensor problems and tools for enabling explorative prototyping of complex sensor problems in the fuzzy front end. Based on the insights, the thesis presents the results as eight hypotheses that may serve as a starting point for further research.

Denne Masteroppgaven omhandler arbeidet som er blitt gjort i forbindelse med utviklingen og implementasjonen av lavkostnads Industri 4.0-prototyper i små og mellomstore bedrifter. Prosjektet har blitt delt i to seksjoner som har blitt utviklet parallelt. Den første seksjonen omhandler utviklingen av et sensorsett som er i stand til å generere data om produksjonen av et trevareprodukt ved overvåkning av verktøy som inngår i prosessen. Den andre seksjonen omhandler utviklingen av brukervennlig programvare som både kan være et læremiddel for brukeren i å sette opp de aktuelle sensorprototypene, samt at det også viser avlesninger fra sensorene til brukeren i sanntid. Denne oppgaven forsøker å svare på tre forskningsspørsmål gjennom prototyping og programvareutvikling. Spørsmålene lyder som følger: 1) Kan data fra lavkostnads Industri 4.0-løsninger nøyaktig vise hvilke prosesser som skjer på en produksjonsbenk for treverk? 2) Er en produksjonsleder med minimal erfaring innen IoT i stand til å sette opp og tolke data fra en Arduino når han eller hun får instruksjoner og ferdig kode? 3) Kan et lavkostnads Industri 4.0-system generere verdifull data for en trevarebedrift der de fleste prosesser er manuelle? Resultatene fra prototypingen er lovende, og har ført til verdifull innsikt som videre kan bidra til forbedringer av det foreslåtte systemet. This Master's Thesis describes the work that has been done in order to develop and implement low-cost Industry 4.0 conceptual prototypes in a small- and medium-sized enterprise. The project has been divided into two main branches which have been worked on in parallel. The first branch concerns the development of sensors which track the assembly of a wooden product by monitoring the tools used in the process. The second branch concerns the development of a user friendly and easy-to-understand software which is capable of both teaching the user to set up the system of sensors on their own, while also serving as an interface which shows the readings from the sensor prototypes in real time. This thesis attempts to answer three research questions through sensor protoyping in parallel with a software development process. The following three research questions are addressed: 1) Can data from low-cost Industry 4.0 solutions accurately represent the activity conducted at a wood product assembly station? 2) Is a production manager with minimal experience with IoT able to set up an Arduino and interpret its data when provided instructions and pre-written code? 3) Does a low-cost Industry 4.0 system generate valuable data for a wood product manufacturer where most processes are done manually? The results from the prototyping activities conducted are promising and have resulted in valuable insight for further improvement of the proposed system.

Insulinpumper er ofte brukt som en del av behandlingen for pasienter med Diabetes Mellitus type 1. Eksisterende forskning viser at jo lavere insulindoser, jo lavere er nøyaktigheten av insulinleveransen. Dette kan være spesielt farlig for pasienter med svært lavt insulinbehov. Det finnes ikke et krav for nøyaktighet av insulinleveranse, og dette er opp til insulinpumpeleverandøren å definere. Leverandørene lover ofte en nøyaktighet på ±5%. I denne avhandlingen skal vi teste nøyaktigheten på insulinleveransen hos forskjellige insulinpumper og diskutere om den er god nok. Denne masteroppgaven er skrevet i samarbeid med Artificial Pancreas Trondheim (APT). APT er en tverrfaglig forskningsgruppe der det langsiktige målet er å utvikle et lukket sløyfe glukose kontrollsystem. De siste årene har det blitt gjort flere forsøk på å utvikle et slikt system av forskjellige forskningsgrupper. Et lukket sløyfe system med automatisk insulinleveranse vil stille enda høyere krav til at det tekniske utstyret er nøyaktig og pålitelig. IEC 60601-2-24 standarden beskriver hvordan insulinpumper skal testes for å verifisere nøyaktigheten og sørge for grunnleggende sikkerhet. I løpet av høsten 2020 implementerte vi denne standarden og etablerte en eksperimentell prosedyre. Prosjektoppgaven kan bli funnet i appendiks D. Denne testmetoden trengte videre utvikling under masteroppgaven for å oppnå tilstrekkelig stabilitet og troverdige resultater. De eksperimentelle resultatene bekrefter at de høye insulindosene er mer nøyaktige enn de lave for insulinpumper fra forskjellige leverandører. Det å følge IEC standarden var deimot utforende fordi den er veldig sårbar mot forstyrrelser. Spesielt når man måler lave doser er den relative feilmarginen betydelig. Derfor bør resultatene bli sett på som en indikator for en nøyaktighetsmåling, og som en mulighet til å sammenligne insulinpumper. Den totale leveransen over tid var nokså nøyaktig for de fleste insulinpumpene, til og med ved lave doser, så når man bruker systemer uten automatiske doseringer av insulin, er ikke avvikene nødvendigvis spesielt klinisk relevante. For lukket sløyfe systemer derimot, krever hver individuelle leveranse et høyt nivå av nøyaktighet. Insulin pumps are commonly used as part of the treatment for patients with Diabetes Mellitus type 1. Existing research indicates that the lower the insulin rates are, the lower the insulin delivery accuracy. Especially patients with low insulin demands might be significantly affected. There is no existing requirement for insulin delivery accuracy; therefore, this is up to the manufacturers to define. Manufacturers often promise accuracy levels of ±5%. In this master thesis, we will test the insulin delivery accuracy of different insulin pumps and discuss whether it is sufficient. This master project is affiliated with Artificial Pancreas Trondheim (APT). APT is an interdisciplinary research group where developing a closed-loop glucose control system is the long-term goal. In the last years, several tries have been made to develop such systems by different researchers. A closed-loop system with automatic insulin delivery will require that the technical equipment is accurate and reliable. The standard IEC 60601-2-24 describes how insulin pumps should be tested to verify accuracy and determine basic safety. During the autumn of 2020, we implemented this standard and established an experimental procedure. The project thesis can be found in appendix D. During this master thesis, the implementation of the IEC standard was further developed to achieve sufficient stability and reliable results. Our experiment results verify that the higher insulin volume deliveries are more accurate than lower ones for insulin pumps from different manufacturers. However, following the IEC standard was challenging because it is very prone to disturbance. Especially when measuring low insulin rates, the relative margins of error are significant. We went through several iterations to optimize the experimental setup. Therefore, we suspect the standard to be poorly suited to test small insulin volumes. Further research should explore alternative methods of measuring insulin delivery accuracy to review the reliability of earlier studies of insulin pump accuracy.

I denne oppgaven undersøkes muligheten for å benytte maskinlæring til å indentifisere glass og metall i kommunalt avfall. Systemet bruker lyd og metalldetektormålinger fra et eksperimentelt oppsett som input for maskinlæringsmodellen. Eksperimentet var konstruert for å simulere tømming av avfall i en avfallsinnsamlingsbil siden denne lokasjonen var det tiltenkte stedet for implementering. Eksperimentet involverte tømming av én pose av gangen, siden det ble ansett som et nødvendig førstesteg. Maskinlæringsmodellen som ble benyttet var en CNN-modell, utviklet iterativt gjennom en ablasjonsstudie. Både multiclass-klassifisering og multilabeling-klassifisering ble testet. Multilabeling var den foretrukne fremgangsmåten, hvor en treffsikkerhet på 96,25% ble oppnådd på uavhengig testdata. Disse resultatene ble oppnådd ved bruk av kondensatormikrofoner, en "Beat-frequency oscillation" metalldetektor, og både Mel-spektrogrammer og MFCC spektrogrammer til å representere lydopptakene. Resultatene viser at systemet fungerer bra for identifisering av glass og metall i individuelle poser, og virker derfor lovende for implentering i den foreslåtte lokasjonen, i tillegg til andre områder. Systemet vil riktignok kreve ytterligere testing etter installasjon på en avfallsbil ettersom det kun ble testet i en eksperimentell sammenheng. In this thesis, the application of machine learning to identify glass and metal in municipal waste is investigated. The system utilizes sound and metal detector data from an experimental setup as inputs for the machine learning model. The experiment was made to simulate the emptying of waste performed by a waste collection truck, as this location was the intended implementation location. The experiment involved the emptying of one bag at a time, as it was seen as a necessary first step. The machine learning model used was a CNN model, developed iteratively through an ablation study. Both multiclass classification and multilabel classification was tested. Multilabeling was the favored approach where an accuracy of 96.25\% was reached on independent test data. These were obtained using condenser microphones, a Beat-frequency oscillation metal detector, and both Mel spectrograms and MFCC spectrograms to represent the sound recordings. The results show that the system works well for identifying glass and metal in singular bags, and thereby shows promise in the proposed location, as well as other implementation areas. The system will however require some additional testing after installation on a collection truck, as it was only tested in the experimental setting.

The two main aims of the thesis are firstly to qualitatively and quantitatively research the effect of using a prototype-driven wayfaring approach for the early-stage development of cyber-physical systems to give normative recommendations to developers and secondly to show and discuss how this kind of research with human subjects can be conducted and improved. Many development methodologies rely on early definition of requirements and design specifications and lack the agility to make fast design adaptations. However, this agility is necessary to cope with unforeseen design problems that arise during the early stages of highly uncertain complex development projects. Prototype-driven approaches, that rely on frequent testing and evaluation of prototypes for advancing the design, offer the necessary agility but research is required to better understand what influences the outcome of such a development process and when it is appropriate to use it. I answer these questions in the context of developing cyber-physical systems by studying two case studies and a controlled quantitative experiment in the context of programming a mobile autonomous robot. The first case study of a small but multidisciplinary project is used to extend the wayfaring approach and shows how prototyping helps to discover and solve unforeseen problems, especially while merging different disciplines. The second case study explains how the wayfaring approach can be applied in highly uncertain large-scale development projects as shown here with the example of the ATLAS detector development at CERN. The controlled experiment quantitatively explores whether early testing of prototypes has an influence on the performance of the design outcome and also explores qualitatively what influences this performance. The research shows that a wayfaring-like approach can be suitable during the explorative concept evaluation phase of solving highly uncertain engineering problems. Testing prototypes and abductively reflecting on the test results helps to discover unforeseen design errors and in finding possible solutions. However, the experiment could not show a statistically significant performance difference between participants who tested their designs early and often compared to participants who could not test their design early. It showed large individual performance differences between participants regardless of their experimental condition. Qualitatively, these individual differences can be explained by variations in error finding ability which not only depends on the participant’s mental ability and programming experience but also on when, how and what the participants test and how the code is written. Furthermore, the results suggest that urged early testing without sufficient questioning of design choices can lead to design fixation supposedly enhanced by the sunk cost of building and testing prototypes. The insights lead to the normative suggestions to use early low-resolution prototyping to keep the sunk costs to a minimum, prototype only the critical functions of several disciplines quasi-simultaneously, prototype different concepts in parallel and prioritize the critical functions whose solutions appear to be least likely to succeed. Well controlled quantitative human subject experiments are necessary to better understand, quantify and compare design methodologies. Such quantitative experiments in design methodology research are rare and often not controlled enough to unambiguously attribute the observed results to the stimulus. This thesis uses the above mentioned robot experiment as a case to show measures like automating the experiment procedures or avoiding direct human-human interaction between the experimenter and the participant to increase repeatability and internal validity of the experiment. The thesis also discusses lessons learned and when using the wayfaring approach is appropriate for developing such experiments.

The world is facing numerous challenges, and entrepreneurship represents one method for introducing solutions to solve these challenges. For this reason, entrepreneurship is more relevant today than ever. Yet when we examine those who want to be entrepreneurs, or generate entrepreneurial outcomes, we see that a great many of them do not succeed. When we dig further we see that there is a cognitive element to this, as nothing physically prevents people from becoming entrepreneurs. We posit that it is an entrepreneurial mindset ("EM") that allows some people to take entrepreneurial actions under circumstances of uncertainty in order to create ventures. This leads to the primary research question of this thesis, which is: what is the role of EM in separating those who take entrepreneurial actions from those who have the intention to take entrepreneurial action but never do? In order to examine this research question, I first began by examining three sub-questions that inform our overarching research question. The first of these is: how do you define and conceptualise what is meant by an EM? In order to answer this, I begin with an overview of key literature on mindset and cognition. Two themes became apparent to me from reading literature on this topic. The first is that there has been thorough research into individual cognitive elements, such as biases, heuristics, beliefs, roles, schemas, identity, expert cognition, meta cognition and personality - yet none of these individually provides insights into what is EM. The second theme which comes through from examining existing literature is that existing conceptions of EM are focussed on what I describe as "high level descriptions": they describe broad outcomes such as a willingness to act under uncertainty, or an ability to identify and exploit opportunities. These conceptions are overly broad and are focussed on the expected outcomes of having an EM, instead of describing what an EM is. To resolve this gap between individual parts of cognition and high level broad descriptions, I set out to provide a conceptualisation of EM. I define EM as: An automated non-conscious perspective including the sum total of cognitive processes; that lead to an individual's willingness to take action under uncertainty, make errors, learn from those failures and direct that learning to specific goals to best solve the entrepreneurial tasks within the process of venture creation (Lynch & Corbett, 2019). This new conceptualisation of EM is based on bringing across work from the field of cognitive psychology and synthesising prior research from outside of entrepreneurship. What separates the proposed conception of EM from prior literature is an acceptance that is not one thing or concept, but rather it is something far more fluid and changing. This conception allows for both mistakes and learning, and is important as a distinction from existing conceptions of EM. Having conceived of what we1 mean by EM, the thesis moves onto examining the second researchquestion: how can we measure EM? We show that prior research measures something other than mindset, and it is often focussed on aspects of personality, despite entrepreneurship as field having moved on from personality as a crucial determining factor in entrepreneurial outcomes. Alternative measures also focus on individual elements of cognition, such as risk propensity, need for achievement or meta-cognition. While these elements all play a role in EM, they only partially contribute to our conception of EM. We therefor set out to test a novel method for measuring EM, using language as a proxy for cognition. This is based on the argument that what we talk about and how we talk about it reflects what we think about it and the way we think about it. Based on this argument, our research examines the ways in which expert entrepreneurs use language differently from less expert entrepreneurs. The result is five themes that are consistent with similar work on measuring how expert entrepreneurs think with regards to venture creation. Having conceived of EM and developed a methodology to measure it, we then set out to examine how to teach EM. This is an important theme based on that bringing about change in the world requires more entrepreneurs to take more entrepreneurial action. This thesis is built around the idea that EM plays a role in giving individuals the mental resources they need to bring ideas into reality. Therefore, finding an appropriate pedagogy that fosters this is crucial. In examining the existing literature on pedagogy, we find that learning from experience has gained popularity, and that within this field, design ("DT") as a practice orientated methodology can be useful for teaching EM. This leads to the third and final research question: why is DT a useful pedagogy for teaching EM? In investigating this research question I first describe our conception of DT, and explain how this contributes to having students practice the different elements of EM. We then delve into studentreflections on DT to elaborate on what students receive out of this pedagogy. In order to provide some practical suggestions on how to teach DT, we also include a section on the practical aspects of organising teaching interventions based on DT. These suggestions are based on prior experience with teaching DT in the context of entrepreneurship. Lastly, in order to answer the overarching research question and explain how EM plays a role in entrepreneurial action, I also describe the way EM influences an entrepreneur's world view, and their choice in responding to entrepreneurial stimuluses, such as opportunities. I show congruence between our conception of EM and existing literature and elaborate on the way EM impacts goal achievement/venture creation. I link EM with a model from cognitive psychology called the Rubicon model. This model shows the stages that people move through in their quest to achieve challenging goals, and the role of mindset in this process. The model highlights the two key mindsets used by entrepreneurs: elaborative mindset and implementive mindset. The thesis is based on 9 individual research articles. These articles use a variety of methods, including qualitative, qualitative, case studies, participant observation, surveys and interviews. The range of articles and methods provide a way to triangulate the perspectives discussed, and each article contributes to the research questions in different ways, with a sense of overlap often existing between the articles. As a body of work, this thesis represents contributions in the form of providing a new conceptualisation of EM and provides a synthesis of existing perspectives. If offers up several new models of the role of EM in influencing behaviour of entrepreneurs. The thesis also provides a new method of measuring EM and provides insight into how to teach EM in a way that has not been fully researched before. Like all pieces of research it also opens many new questions as well as providing new insights on old questions. This thesis plays a role in furthering our insight in the role of EM in creating entrepreneurial outcomes.

Proteser har utviklet seg kraftig de siste tiårene noe som har økt livskvaliteten til protesebrukere. Trøndelag Ortopediske Verksted (TOV) er hovedaktøren for konstruksjon av proteser for protesebrukere i Midt-Norge. De har en lidenskap til å forbedre livet til brukerne sine og ønsker å vise dem at «alt er mulig» selv etter en amputasjon. I vintersesongen har en gruppe fra TOV tatt med brukerne sine ut i alpinbakken. For enkelte, særlig de med låramputasjoner, var dette mer en skremmende opplevelse enn å være givende. Brukerne ble satt i bakken med vanlige alpinski med deres ordinære proteser. I første leksjon, ploging, mistet brukerne totalt kontroll over skien på grunn av glidning i protesehylsen rundt stumpen, den gjenstående delen av beinet. Med sitt driv om å bedre livskvaliteten til sine brukere, ønsker TOV å finne løsninger på dette rotasjonsproblemet. I denne masteroppgaven har nå-situasjonen blitt studert, inkludert anatomi og låramputasjoner. Protesen TOV bruker er en av standardproteseløsningene som brukes i verden i dag. I behovsanalysene som ble gjennomført viste det seg at det er en rekke begrensinger, hva gjelder modifikasjon av protesehylsen og sett fra anatomiens side. En del av målene for denne oppgaven var å kartlegge behovene, både for pasienten som skal bruke protesen, men også for videre utvikling. TOV ser på dette som et langsiktig prosjekt, både hos dem og eventuelt nye prosjekt og masteroppgaver. I kartleggingen kom det også frem at TOV ikke har noen muligheter til å teste protesene sine før de prøver dem ut på pasientene. Mulighetene for resultatavhenging forbedring, er avhengig av å ha en måte å måle utfallene på. Hovedmålsettingen med denne masteroppgaven ble derfor å studere og utvikle måter å teste protesehylser på. Utfallet av denne masteroppgaven inkluderer, i tillegg til bruker og behovsanalysene, en testrigg som er ment til bruk av ortopediteknikere og -ingeniører hos TOV og i videre utviklingsprosjekter I testriggen er det implementert sensorer, som kan gi et inntrykk av hvordan protesen føles på stumpen til en amputert når det blir påsatt en kraft som skal simulere skien i alpinbakken. Prostheses have evolved greatly over the recent years, improving the life of amputees greatly. Trøndelag Ortopediske Verksted (TOV) are making most of the prosthetic devices used by the amputees in the central part of Norway. With passion, they wish to activate their prosthetic users, showing them that it is possible to do great things even after loosing a limb. Alpine skiing is one of these activities. However, when the above the knee amputees were equipped with alpine skis using their ordinary prostheses they lost all control. Skiing got more terrifying than rewarding. The socket, which is made to secure the prosthesis to the residual limb was slipping, rotating around the limb. As TOV wish to improve the life of their users, they want to find solutions to this problem. This research has investigated the initial situation, including anatomy of lower limb amputation, prosthetics and socket design as well as development possibilities of enhancing the socket design. In the research, a need for testing sockets without being dependent on the amputee was emphasized and sat as the main objective. The research and development process continued on to explore possibilities of testing rotation in prosthetic sockets. Using various prototypes, mechatronics including force sensors like load cells and flexible force sensitive resistors was explored and implemented in a test rig. The final testing rig, includes a wooden residual limb model, load cells, force sensitive resistors and an user interface intended for usage by TOV or in a following student project.

Entrepreneurs are an important source of innovation for transformation of the industrial landscape and thus future sustainable economic growth. The relevance for top policy makers is the great effort and large investments used to build entrepreneurship support programmes (ESPs) supporting entrepreneurship projects (EPs). The ESPs consist of education, grants, mentoring, incubation, and investments. There is a lack of empirical evidence on the activities in EPs and on how the ESPs impact EPs in a positive way. For traditional economic ventures, a major source of such empirical evidence comes from analysing the data gathered in companies' management information systems (MIS). This is, however, rarely an option for entrepreneurial ventures, as these usually do not have the resources to have a MIS. The main scientific contribution of this PhD is to offer a way to gather such empirical data on the activities in EPs and on the effect of ESPs on EPs. The way in which this will be achieved is to combine a selection of entrepreneurship theory with practical experience to develop an entrepreneurship activity system (EAS), which again is used to build a specialised entrepreneurship management system (EMS). The EMS is then made available to as many ESPs and EPs as possible. The data from the entrepreneurship management system will then be loaded into an analytics database for hybrid (quantitative and qualitative) analysis. The data will be used for a set of empirical papers casting light on different entrepreneurial sub¬processes. These data will again be used to improve the EAS, the EMS and the selection of theories in a circle of iterative plan-do-check-act cycles. The thesis is built around two research questions: (RQ1) Flow does the EMS need to be designed to secure adaptation by both EPs and ESPs and allow for effective analysis of their activity data?; and (RQ2) To what degree does the EMS data provide observation of EP activity and the impact of the ESPs? Furthermore, the thesis has four research objectives: (RO1) design an entrepreneurship management system (EMS) for describing and capturing activities within EPs and between EPs and ESPs in their whole granularity; (RO2) populate that system with a significant sample size of real entrepreneurial activity, ranging over a significant number of ESPs; (RO3) identify potential research methods and strategies deployable using this data; and (RO4) make a proof of concept for the individual methods. Since the fall of 2013,1 have been the main force behind developing such an EMS, including an activity system containing seven steps and 28 elements. I have also been principally involved in optimizing the EMS data for research purposes. Since its introduction, different developmental versions of the EMS have been introduced to more than 17.000 EPs. There has been collaboration with more than 30 incubators, universities, banks, venture capitalists, and other ESPs in distributing the EMS. Based on the experiences and data gathered, the nine papers (C1-C9) that are the bases for this PhD were written. Eight of them has been published while the last (C2) is currently pending review. The main conclusions of the thesis are that an EMS can be built in such a way that it provides value to EPs and ESPs and thus will be used by them. To be valuable without the support of an ESP, the EMS should contain a massive online open course (MOOC) covering both the tutorial on how to use the EMS and the entrepreneurial theory supporting it. Another conclusion is that an EMS can provide observation on EP activity and the impact of ESPs. The final contribution will be the distribution of the production version of the EMS starting 2020. The main effort in utilising the EMS has yet to come, as it will be made generally available for research purposes.

40% av de 81 deltagende kommunene taper mer en 40% av deres drikkevann til distribusjonsnettet deres, i følge rapporten Norsk Vann (2017). Kun 50% av de studerte distribusjonsnettene hadde en tilfredsstillende standard. Lekkasjer kan føre til trykkfall, som kan forårsake at distribusjonsnettet blir forurenset. Forurenset drikkevann kan være farlig og medføre sykdom og, til og med, død. Dette er ikke bærekraftig med helsemessige, miljømessige eller økonomiske hensyn. Hensikten med denne masteroppgaven er å utvikle et nytt lavkost-konsept for å lokalisere vannlekkasjer i rør, ved å bruke Wayfaring Method i et rammeverk av Engineering Design. Utforskende prototypebygging med raske design iterasjoner har blitt brukt for å undersøke ideer med design-build-test sykluser. På denne måten har kunnskapshull blitt tettet og oppdukkende behov for løsningen blitt definert. Fra et før-prosjekt hvor flere lavkost måter å måle avstander under vann ble utforsket, ble det uttenkt et konsept som kunne brukes til å lokalisere rørlekkasjer. Konseptet lar en passiv sonar flyte med vannstrømmen i et rør, og måler avstanden tilbakelagt med en pinger posisjonert på et kjent punkt. I dette konseptet har to nøkkelvariabler blitt identifisert, som tilsammen bestemmer lokasjonen til en lekkasje. Disse variablene er avstanden mellom pingeren og sonaren, og lydtrykket rundt sonaren. Tre delsystem av prototypen ble designet, bygget og testet parallelt til å møte sine individuelle krav satt fra starten, samt krav oppdaget under prosessen. Da de individuelle kravene ble møtt, ble delsystemene satt sammen og testet i mer relevante testmiljø. Den siste prototypen ble testet til å være i stand til å måle de to nøkkelvariablene på et tilstrekkelig vis. Selv om avstandsmålingen var nøyaktig i et vannbassengmiljø med en mer en tilstrekkelig ekstrapolert rekkevidde, var den ikke i stand til å gjøre tilsvarende målinger da plassert i det faktiske rørnettet. Dette avslørte ikke bare utilstrekkelig kunnskap om problemrommet men også utilstrekkelig kunnskap om hvordan prototypen samhandler med den tiltenkte miljøet. Derfor blir det foreslått at produktutviklere bør fra tidlig fase aktivt alternere mellom å teste prototypen i dens tiltenkte miljø og konstruert testmiljø, i fremtidig arbeid. I forbindelse med prosjektet ble det sent inn en vitenskapelig artikkel til NordDesign 2020 konferansen, som omhandlet dette. Basert på resultatene blir konseptet vurdert som gjennomførbart, men med viktige ubesvarte spørsmål. Videre blir det foreslått et sett med spesifikke endringer på prototypen som kan øke rekkevidden, samt andre forbedringer som kan legge til kapasiteter og øke konseptets bruksområde. Utfallet av dette prosjektet er en fungerende prototype av en ny lavkost løsning med lovende resultater, og et behov for videre utvikling. 40 % of the 81 participating Norwegian municipalities lose more than 40 % of their freshwater to their water distribution network, according to Norsk Vann (2017). Only 50% of the studied network had a sufficient standard. Leakages may cause drops in pressure, which may release contaminants into the distribution network. Contaminated water can be dangerous and may cause sickness and even death. This is not sustainable in terms of health, environment or economy. This master thesis aims to develop a new low-cost concept to locate water pipe leakages through the use of the Wayfaring Method within a framework of engineering design. Exploratory prototyping with rapid design iterations has been used to probe ideas with design-build-test cycles. In this way, knowledge gaps have been bridged, and emerging requirements for the solution have been defined. From a pre-project where several ways of measuring distances underwater in low-cost manners were explored, a concept that could be applied in locating leakages emerged. The concept lets a passive sonar flow with the water stream in a pipe, and measures the distance traveled with a pinger positioned at a known point. With this concept, two key variables have been identified, which together determine the location of a leakage. These are the distance between the pinger and the sonar, and the sound pressure level (SPL) ambient to the sonar. Three subsystems of the prototype were designed, built, and tested concurrently to meet individual requirements set from the start, as well as emerging requirements. When the individual requirements were met, the subsystems were integrated into more relevant test environments. The final prototype was tested to measure the two key variables successfully. Although the distance measurement was accurate and had a sufficient extrapolated range in an underwater environment, the results were not equal when tested in an actual pipe system. This unveiled not only insufficient knowledge of the problem space but also insufficient knowledge of how the integrated prototype interacts with its intended environment. Therefore, it is suggested that the product developer should early-on actively alternate testing the prototype integrated into the actual intended environment and prototyped test environments in future work. During the project a paper concerning this experience was submitted to the NordDesign 2020 conference. Based on the results of the project, the concept is considered feasible, with important unanswered questions. Furthermore, it is suggested a set of specific alterations on the prototype that may increase range, as well as other improvements that may add capabilities and increase the concept’s application field. The outcome of this project is a prototype of a new low-cost solution with promising results, and a need for further development.

Denne avhandlingen er adressert til forskere innen produktutvikling. Den har som mål å styrke og forbedre produktutviklingsforskning ved å tilby bedre metoder og verktøy for å forske på prototyping i prosjekter innen tidligfase produktutvikling. Avhandlingen argumenterer for at det er et behov for en ny metode som muliggjør at forskere kan gjøre flere observasjoner fra utviklingsprosjekter, uten å bruke så mye ressurser som eksisterende forskningsmetoder krever, og at dette behovet kommer fra begrensningene til metodene, verktøyene og ressursene som er tilgjengelige for produktutviklingsforskeren. Hovedbidraget i avhandlingen er derfor å presentere enmetode for a samle fysiske prototyper fra aktive prosjekter innen tidligfase produktutvikling, hvilket muliggjør analyse av store datasett med prototyper. Avhandlingen hevder at å samle fysiske prototyper er en gjennomførbar løsning for å samle større datasett fra aktive produktutviklingsprosjekter, da dette krever betraktelig mindre arbeidsinnsats fra forskeren sammenlignet med eksisterende forskningsmetoder. For å teste og evaluere den foreslåtte metoden har et system for digital innsamling av fysiske prototyper blitt utviklet. Dette systemet har som mål å møte et sett med funksjonskrav som har blitt identifisert som viktige for å implementere den foreslåtte metoden. Systemet har blitt utplassert på to forskjellige lokasjoner; i produktutviklingsavdelingen til en større norsk bedrift og i forskningslaboratoriet TrollLABS på NTNU. Et datasett på over 950 fysiske prototyper har blitt samlet inn digitalt gjennom bilder og metadata i løpet av dette prosjektet—noe som demonstrerer at den foreslåtte metoden er gjennomførbar for å samle forskningsdata fra aktive tidligfase produktutviklingsprosjekter. Avhandlingen argumenterer for at den foreslåtte metoden kan brukes til både kvantitative og kvalitative undersøkelser av tidligfase produktutviklingsprosjekter, og viser hvordan dette kan gjøres ved innsamling av prototyper fra ett prosjekt. En utfordring som oppstår når man samler større datasett med fysiske prototyper er at det kreves store ressurser for å analysere datasettet. Denne avhandlingen viser flere mulige løsninger på dette problemet ved å automatisk klassifisere ulike egenskaper fra bilder av prototyper gjennom å trene om eksisterende modeller med spesialtilpassede datasett—noe som viser at slike løsninger kan redusere ressursene som kreves for å analysere prototyper betydelig.

Denne masteroppgaven tar for seg tidlig-fasen av et utviklingprosjekt for en mekanisk lunge, for en pasient simulator. Prosjektet er utført på TrollLabs på NTNU, og oppgaven er gitt av Laerdal Medical. De had har et behov for en lungsimulator som er takler å ventileres kunstig, fra en respirator, og som kan plasseres inne i en av deres avanserte pasient simulatorer. Prosjektet eksisterer i en tidlig fase, hvor kravspesifikasjoner ennå ikke er satt. Prototyping er brukt som et viktig verktøy for å generere kunnslap. Dette inkluderer en tidlig fase av hurtig prototyping, som holder et høyt tempo gjennom iterative læringskurvere basert på en brukersentrert holdning til produltdesign og utvikling. Prototypingen er var rettet mot a gjennskape akseptable, menneskelige lungevolum og trykk, gjennom en utforskning av forskjellige materialer og geomtriske former for den mekaniske lungen. Kravspesifikasjonene utviklet seg mot en lunge med variabel stivhet og med muligheten til å etterligne egenksapene til lunger med ulike karakteristikker og sykdommer. Variabel luftveismotstand har også vært en viktig faktor, i tillegg til passient triggering av assisterte pustemoduser på ventilatoren. Denne oppgaven kommer med noen forslag på metoder for hvordan lungen skal forholde seg til positivt ende-ekspiratorisk trykk (PEEP) og variabel lunge compliance. De ulike funksjonene ble satt sammen til en modulær prototype med visse variable kvaliteter, og ble testet med en respirator med assistanse fra en doktor i intensiv medisin. Det foreslåtte konseptet skal representere et skritt i rettningen mot et ferdig prodult, men beffinner seg fortsatt i et tidlig utviklingsstadie. Videre arbeid må til for å bedømme kvaliteten og virkningen av de foreslåtte løsningene og funksjonskravene. Dette burde gjøres testing med høyere oppløselige prototyper samt å innvolvere flere burkere i prosessen. This master's thesis describes the early stage development project of a mechanical lung for a patient simulator. The project is done at TrollLabs (NTNU) and the task is provided by Laerdal Medical. They have requested a lung simulator that is capable of being ventilated with artificial respiration from a ventilator and that can be placed inside one of their advanced patient simulators. The project exists in an early pre-requirement phase of product development, so prototyping has been a main tool for gathering knowledge. This includes a period of rapid prototyping and fast paced, iterative learning from a user-centred design approach. The prototyping effort was aimed at recreating acceptable human lung volumes and pressures, thorough an investigation of materials and geometric shapes for the mechanical lung. The requirements developed into a focus on variable lung compliance (stiffness) to be able to replicate the properties of lungs with different characteristics or diseases. Variable airway resistance has also been an important factor, as well as patient triggering of assisted ventilation modes. The thesis suggests some methods for managing different positive end expiratory pressures (PEEP) and variable lung compliance. The different functions were put together in a modular prototype with certain variable qualities, that were tested by a ventilator with the assistance of a doctor in intensive care medicine. The proposed concept represents a direction towards a final product, but is still in a very early stage of development. Future work may determine the efficacy of the proposed solutions and functional requirements by doing higher resolution tests and including more users in the process.

I 2016 utgjorde hjerte- og karsykdommer 31% av alle dødsfall på verdensbasis, og hypertensjon er blant de viktigste årsakene til disse. Det har derfor blitt forfektet at høyt blodtrykk skal vurderes som et problem av dimensjoner, og ikke bare en sykdom. Samtidig anslås det at nesten alle tidlige dødsfall forårsaket av hjerte- og karsykdommer kunne vært forhindret ved bedre fysiologisk monitorering. Dette gjør forbedringer av dagens fysiologiske sensorer til en vital del i bedringen av helse rundt om i verden. Til tross for dette er de fleste av dagens blodtrykksmålere basert på en 100 år gammel teknikk, og brukervennlighet regnes som den viktigste faktoren for at den fremdeles ikke er utbredt utenfor legens kontor. Dette prosjektet tar seg dermed som mål om å utvikle en blodtrykksmåler som er behagelig og praktisk, samtidig som den estimerer blodtrykk nøyaktig og ofte. I utviklingsprosessen er både ny og etablert forskning undersøkt og analysert. Gjennom dette har detektering av fysiologiske responser fra den pulserende radiale arterien og huden pekt seg ut som interessant av både fysiologiske og historiske grunner, samtidig som det gir muligheter for å undersøke lite utforskede blodtrykksrelasjoner. I utviklingen av sensorløsningen er ’wayfaring-modellen’ brukt, og en rekke ulike fysiologiske sensorer er designet, utviklet og testet. Den praktiske erfraringen og de empiriske bevisene fra de ulike prototypene har sammen med relatert teori vist at flere av de testede prinsippene komplementerer hverandre gjennom deteksjon av separate mekanismer beslektet til blodtrykk. En miniatyrisert sensor med mulighet for m°aling av fotopletysmografi, tonometri, ballistokardiografi, elektrodermal aktivitet, hudtemperatur og puls CO-oksimetri er derfor utviklet. Basert på innledende brukertesting har sensoren vist seg å være både komfortabel og praktisk i bruk. Nøyaktigheten og muligheten for kontinuerlige målinger er derimot ikke testet. Dette utgjør dermed en naturlig fortsettelse av prosjektet, hvor kvantifisering av hvordan de ulike sensorelementene påvirker blodtrykksestimater, samt identifikasjon av hvilke parametere som er uviktig, står sentralt. Selv om den endelige prototypen ikke er fullstendig testet diskuteres ulike aspekter rundt dens nøyaktighet, i tillegg til potensielle forbedringer som kan vurderes for fremtidige designiterasjoner. In 2016, cardiovascular diseases represented 31% of all deaths worldwide. Hypertension is considered one of the main contributors to the burden of cardiovascular diseases, and it has been championed that an elevated blood pressure should not be considered a disease, but a problem of magnitude. At the same time, it is estimated that almost all premature deaths caused by cardiovascular diseases are preventable by improved physiological monitoring. This makes better blood pressure monitoring a major target for improved health and wellbeing worldwide. However, the cuff-based measurement technique introduced more than 100 years ago is still the mainstay of today’s measurement devices, and its usability is viewed as a major factor impeding user acceptance outside the physician’s office. This project is therefore focused on the design and development of a sensor system for wearable blood pressure monitoring that is both comfortable and convenient, in addition to ensuring accurate blood pressure predictions and allowing for frequent measurements. To do so, a broad range of novel and established blood pressure-related research are explored and assessed. Based on this, the detection of physiological responses from the arterial pulse and the skin above the radial artery is found to be advantageous for both physiological and historical reasons, and, at the same time, enabling for investigation of understudied blood pressure-related mechanisms. In the development of the sensor solution, the ’wayfaring model’ is employed, and a number of different physiological sensors are designed, developed, and tested. From the practical experience and empirical evidence gained from the prototypes and their related literature, it is found that many of the detected responses contain information of separate physiological mechanisms linked to blood pressure, and thus complement each other well. A miniaturized sensor solution comprising photoplethysmography, tonometry, ballistocardiography, electrodermal activity, skin temperature, and pulse CO-oximetry is therefore developed, and preliminary user tests suggest that the final prototype is both convenient and comfortable. Its accuracy and potential for frequent measurements are, however, yet to be validated. Validation, therefore, forms a natural starting point in the continuation of the project, and it is considered to be indispensable to quantify how the different sensor elements affect the precision of estimates as well as to identify which of the parameters that do not have significant effects on the predictions. However, although the final prototype is yet to be validated, different aspects regarding its accuracy are discussed, together with potential improvements for future design iterations.

Denne avhandlingen er adressert til forskere innen produktutvikling. Den har som mål å styrke og forbedre produktutviklingsforskning ved å tilby bedre metoder og verktøy for å forske på prototyping i prosjekter innen tidligfase produktutvikling. Avhandlingen argumenterer for at det er et behov for en ny metode som muliggjør at forskere kan gjøre flere observasjoner fra utviklingsprosjekter, uten å bruke så mye ressurser som eksisterende forskningsmetoder krever, og at dette behovet kommer fra begrensningene til metodene, verktøyene og ressursene som er tilgjengelige for produktutviklingsforskeren. Hovedbidraget i avhandlingen er derfor å presentere enmetode for a samle fysiske prototyper fra aktive prosjekter innen tidligfase produktutvikling, hvilket muliggjør analyse av store datasett med prototyper. Avhandlingen hevder at å samle fysiske prototyper er en gjennomførbar løsning for å samle større datasett fra aktive produktutviklingsprosjekter, da dette krever betraktelig mindre arbeidsinnsats fra forskeren sammenlignet med eksisterende forskningsmetoder. For å teste og evaluere den foreslåtte metoden har et system for digital innsamling av fysiske prototyper blitt utviklet. Dette systemet har som mål å møte et sett med funksjonskrav som har blitt identifisert som viktige for å implementere den foreslåtte metoden. Systemet har blitt utplassert på to forskjellige lokasjoner; i produktutviklingsavdelingen til en større norsk bedrift og i forskningslaboratoriet TrollLABS på NTNU. Et datasett på over 950 fysiske prototyper har blitt samlet inn digitalt gjennom bilder og metadata i løpet av dette prosjektet—noe som demonstrerer at den foreslåtte metoden er gjennomførbar for å samle forskningsdata fra aktive tidligfase produktutviklingsprosjekter. Avhandlingen argumenterer for at den foreslåtte metoden kan brukes til både kvantitative og kvalitative undersøkelser av tidligfase produktutviklingsprosjekter, og viser hvordan dette kan gjøres ved innsamling av prototyper fra ett prosjekt. En utfordring som oppstår når man samler større datasett med fysiske prototyper er at det kreves store ressurser for å analysere datasettet. Denne avhandlingen viser flere mulige løsninger på dette problemet ved å automatisk klassifisere ulike egenskaper fra bilder av prototyper gjennom å trene om eksisterende modeller med spesialtilpassede datasett—noe som viser at slike løsninger kan redusere ressursene som kreves for å analysere prototyper betydelig.

Denne masteroppgaven beskriver tidligfaseutviklingen av en legemiddelbeholder for rask oppløsning av Acetylsalisylsyre(ASA). ASA er en kjent blodfortynner og er et etablert førstehjelpstiltak når det er mistanke om hjerteinfarkt. Raskt opptak av ASA er derfor ønskelig for å mitigere risikoen for et dødelig utfall av hjerteinfarktet. Asamedic har utviklet et nytt løsemiddel for ASA som gir høy løslighet av ASA, og å maksimere oppløsningshastigheten er derfor ønskelig. Denne oppgaven identifiserer utfordringer knyttet til oppløsningsprosessen og beskriver videre produktkrav avledet fra prototyping og testing. Forfatteren har bygget og testet 22 prototyper for å adressere de ulike utfordringene til oppløsningsprosessen for å maksimere oppløsningsraten til ASA Denne oppgaven bruker den utforskende designmetoden Wayfaring. Denne metoden tillater en fleksibel design tilnærming for å legge til rette for raske iterasjonssykluser til å teste prototyper og konsepter. En introduksjon til de brukte metodene er presentert for å forstå valgene som er gjort i denne oppgaven. Det foreslåtte konseptet tar hånd om de identifiserte utfordringene for oppløsning av ASA som er deaggregering, sikret fukting, og agitasjon. Oppvarming har blitt vist å forbedre oppløsningen og et konsept for oppvarming er lagt til i det foreslåtte konseptet. Konseptet som er presentert er testet og sammenlignet med den nåværende løsningnen og har vist en 81% forbedring i oppløsningstid. En referransemålstabell er laget for å belyse styrker og svakeheter til hver prototype. Det presenterte konseptet er en tidligfasemodell av en legemiddelbeholder som skal brukes til raske oppløsningsprosesser. Effekten av konseptet har blitt sammenlignet med referansemål på tilsvarende løsninger og utviklede konsept. Konseptet trenger fortsatt mer designarbeid for å gjøres klar til produksjon. Fremtidig arbeid er nødvendig for å måle nøyaktig oppløsningsrate til ASA. This master's thesis describes the early-phase development of a pharmaceutical container for the rapid dissolution of Acetylsalicylic acid (ASA). ASA is a known thrombocyte inhibitor and is an established first aid measure taken when Acute Myocardial Infarction (AMI) is suspected. Rapid onset of ASA is therefore highly wanted to mitigate the risk of a deadly outcome of the AMI. Asamedic has developed a novel solvent for ASA that allows for high solubility of ASA, and to maximize the dissolution rate is therefore wanted. The thesis identifies the challenges of the dissolution process and describes it according to the requirements derived from prototyping and testing. The author built and tested 22 prototypes to address the different challenges of the dissolution process to maximize the dissolution rate for ASA. This thesis makes use of the explorative design method Wayfaring. This method allowed for a flexible design approach to aid fast iteration cycles to test prototypes and concepts. An introduction to the used methods has been included to aid the understanding of the choices made in the thesis. The proposed final concept tackles the identified challenges of the dissolution of ASA, namely deaggregation, assured wetting, and agitation. Heating has been shown to improve the dissolution process and a concept for heating is included in the final concept proposal. The proposed concept has been tested and compared to the current solution to the challenge and shows an 81% reduction in dissolution time. A benchmarking table is included to compare different prototypes and highlight the strengths and weaknesses of each prototype. The presented prototype is an early concept model of a pharmaceutical container to be used for rapid dissolution purposes. The efficacy of the prototype has been benchmarked against current solutions and developed concepts. The concept still needs more design work to be ready for manufacturing. Future research is needed to accurately measure the dissolution rate of ASA.

Denne masteroppgaven er skrevet i samarbeid med Alcoa Mosjøen og omhandler tidligfase konseptuell testing av mulige inspeksjonsmetoder for å finne sprekker i karbon anoder. Anodene er brukt som offeranoder i Hall-Héroult prosessen. Sprekker er et kjent problem i anodeproduksjonen, og kan lede til feil i elektrolyseprosessen og økonomiske belastninger. Inspeksjon av sprekker i anoder er i dag gjort manuelt ved at en ansatt visuelt inspiserer et utvalg av anodene i produksjonen. En prosjektoppgave som omhandlet samme tema ble skrevet høsten 2018, hvor mulige in- speksjonsmetoder for sprekker ble undersøkt. Dette forarbeidet har vært et grunnlag for videre arbeid i denne masteroppgave hvor et utvalg av metoder blir undersøkt i større detalj. Oppskaleringspotensialet til computed tomography ble utprøvd med både medisinske- og industrielle systemer. Ingen av testene ga tydelige indikasjoner på at de kan skanne en full-skala anode. En spesialtilpasning av teknologien vil trolig gi gode resultater likevel. Termografi viste seg å være en håpefull kandidat for å detektere overflatesprekker basert på tester i realistiske omstendigheter ved Alcoa Mosjøen. Radar og måling av tetthet som indikasjon på sprekker var nye inspeksjonsmetoder som be undersøkt i denne oppgaven med mangelfulle resultater. This Master’s thesis is written in collaboration with Alcoa Mosjøen and concerns early-stage conceptual testing of inspection methods suited for finding cracks in carbon anodes. The anodes are used as the sacrificial end of the Hall-Héroult process. Cracks are a known problem during the anode production, leading to faulty operation of the electrolysis and financial losses. Inspection for cracks in anodes are to this date done manually by having an employee visually inspect samples from the production. A project thesis has done preliminary investigation of possible crack detection methods. Methods discussed in this thesis are based on those results but are investigated to a greater extent. Computed tomography had its scale-up capabilities tested with both medical- and industrial commercial systems. Neither of these tests resulted in unambiguous indications that a computed tomography system can radiate a full-scale anode. Regardless, a custom adaptation of this technology can likely give good results. Thermal imaging proved a hopeful candidate for finding surface cracks, having reassuring results from high-fidelity tests at Alcoa Mosjøen. Radar and density tests as an indication for crack formation were new inspection methods tested in this thesis with unfavourable results.

Hjertestans utenfor sykehus rammer årlig rundt 3,000 mennesker i Norge. Dødeligheten er høy og stiger dess lengre tid som går før pasienten får behandling. Et eksperimentelt nødinngrep, REBOA, vurderes som en mulig behandling i felt, utført av luftambulansepersonell. Dette personellet må trenes. TrollLabs ble kontaktet for å utvikle et vaskulært fantom for å tilrettelegge trening av leger og redningsmenn i en pilotstudie. Fantomet måtte være ultralydkompatibelt, og et spesialisert kateter måtte kunne plasseres. Gjennom en utforskende prototypingsprosess ble dette oppnådd med en ny kombinasjon av materialer. Et lag parafinoljegele, omringende simulerte blodkar, er innesluttet i et erme av silikongummi. Resultatet er en komposittstruktur som gir en veldig realistisk vevsmotstand, og et meget bra ultralydbilde. Rundt 30 fantomer ble levert til prosjektinitiatorene og samarbeidspartnere. Flere av disse ble brukt i et treningsprogram ved medisinsk simulatorsenter ved St. Olavs hospital. 7 anestesileger og operatører ved prehospital akuttmedisinsk tjeneste fra Luftambulansetjenesten deltok. Dette treningsprogrammet er beskrevet i detalj i to publikasjoner som stammer fra bidragene i dette prosjektet (BMJ Open, 2019; Circulation, 2018). Hardhetstesting ble utført for å evaluere materialoppførsel, og lekkasjeraten ble målt under tre forskjellige brukstilstander for å evaluere levetid. Den leverte løsningen ble ansett av brukere som like god eller bedre enn kommersielt tilgjengelige alternativer. Prosjektet mottok opp mot 1 million kroner i støtte gjennom NTNU Discovery og Helse Midt-Norge Innovasjonsmidler for å støtte videre utvikling. En patenteringsprosess for åndsverket er iverksatt. Out-of-hospital cardiac arrest affects around 3,000 persons in Norway every year. The lethality rate is high and increases the longer a patient goes without treatment. An experimental emergency procedure, REBOA, is being considered as a potential treatment in the field, to be carried out by ambulance helicopter crews. These crews must be trained. TrollLabs was approached to develop a vascular phantom to facilitate training of the physicians and paramedics taking part in a pilot study. The phantom needed to be ultrasound compatible, and a specialized catheter had to be able to be inserted. Through a process of exploratory prototyping, this was achieved with a novel combination of materials. A layer of hydrocarbon gel encasing simulated blood vessels is embedded in a silicone sleeve. The resulting composite structure offers a highly realistic level of tissue resistance, and an excellent ultrasound image. Approximately 30 phantoms were delivered to the project initiators and co-operators. A number of these were used in a training programme carried out at the medical simulation centre at St. Olav’s University Hospital. 7 anaesthesiologists and prehospital EMS operators from the National Air Ambulance foundation took part. This training programme is described in detail in two publications stemming from the contributions of this project (BMJ Open, 2019; Circulation, 2018). Indentation testing was carried out to evaluate material behaviour, and the leak rate was measured under three different operating regimes to evaluate durability. The delivered solution was considered by users to be as good as or superior to commercially available alternatives. The project received approximately 1 million NOK in funding to support additional development through grants from NTNU Discovery and Helse Midt-Norge Innovasjonsmidler. A patenting process for the IP has been initiated.

This thesis presents the conceptualization of a new solid-state process for additive manufacturing (AM) of extrudable metals and alloys. The process is based on the Hybrid Metal Extrusion & Bonding (HYB) technology and has been termed HYB-AM. The main objective of this thesis has been to develop the HYB-AM process through concept evaluation and full-scale experimental testing, hence making it possible to assess the mechanical integrity of a deposited structure and understand the governing mechanisms of the process. A novel set-based approach for concept evaluation has been developed and applied. The combination of rapid prototyping and physical modelling made it possible to test and evaluate multiple extruder and deposition concepts without having to build full-scale steel prototypes. Plastic models of the different extruder concepts were produced by rapid prototyping and attached to a small CNC-machine. To test the function of each design, plasticine was processed through the different extruders and deposited on the machine bed. Through the evaluation of all the modelled concepts the extruder design for full-scale testing was identified. Full-scale experiments have been carried out for two different extruder designs. Through the fabrication of samples and the subsequent assessment of the layer bonding, the technical feasibility of the HYB-AM process has been conclusively demonstrated in producing mechanically sound 3D structures. The results from the mechanical testing display an ultimate tensile strength across the bonded layers approaching that of the substrate material, yet at lower elongation prior to fracture. Moreover, inspections of the fracture surfaces show evidence of extensive dimple formation. This indicates that metallic bonding is achieved between the layers. However, regions of kissing-bonds and lack of bonding are also present, thus calling for further process optimization. A novel method for fabrication of tensile specimens for assessing the bond strength between the layers has been developed and applied. The miniature specimens are machined from the samples using a single thread milling tool. In this way, the specimens can be located such that the interface between separate layers crosses the reduced section of the specimens. This allows the bond strength between the layers to be tested in pure tension. FEA has been used to study material flow in the extruder, as well as the conditions at the interfaces of the deposited extrudate and the substrate in order to identify and characterize the process parameters involved. Analysis of the material flow, shows that the extrusion pressure is virtually independent of the deposition rate. Furthermore, from the simulations of the material deposition sequence it is visible how the contact pressure at the interface will drop below the bonding threshold if the feed speed becomes too high relative to the material flow through the die. The results obtained in this work, will be useful in the further development towards industrialization of the process. Although this work has demonstrated the process in full scale and established an understanding of the process, further research is required prior to industrialization.

Det ble ytret et ønske om å utvikle en innretning som kunne brukes til å måle mekaniske egenskaper av en ski, primær bøyestivhet, men muligheter for testing av torsjonsstivhet sees på som fordelaktig. Innretningen skal brukes som et verktøy for å finne eventuelle problem områder for å gjøre utviklingen av nye og bedre produkter mereffektive. Målet er at test riggen også kan brukes som et verktøy for kvalitetskontroll mellom leveranser fra leverandøren. Gjennom prosjektet har tidligere forskning, brukere opplevelse, og bedriftens krav og spesifikasjon har fungert som retningslinjer for å etablere en visjon om hva som skal utvikles. Testriggen er konstruert i henhold til disse krav og spesifikasjoner. Ulike belastningsscenarier og sensorer for måling av avbøyning og vridning av ski ble testet gjennom hele prosjektet. Dette for å observere best mulig løsning tilpasset forhånds definerte behov og spesifikasjoner, samt å øke innsikt og læring. Konsepter og belastningsscenarier har blitt evaluert ved hjelp av evalueringsmatriser og utførte tester. Test-riggen og sensorenhetene er bygget som følge av dette. Bøyestivhet og torsjonsstivhet har blitt beregnet ut fra målinger som er gjort ved hjelp av sensorene. Arduino har blitt brukt til å programmere de forskjellige sensorene som ble brukt gjennom prosessen. Python har blitt brukt til å behandle data samlet av sensorenhetene. Det har blitt utført tester for å måle bøyestivhet og torsjonsstivhet, og tester for å verifisere enhetens repeterbarhet Effekten av montert binding og skistøvel på skiens bøyestivhet har blitt studert, basert på resultatene oppnådd av F.Rainer and Kaps (2004), De Gobbi and Petrone (2008) og Federolf et al. (2006) A wish to make even better skis has led to a desire to develop a device to measure the mechanical properties of a ski, primarily bending stiffness, and opportunities for testing torsional stiffness was also considered advantageous. The test-rig will serve as a tool to locate any problem areas to make the development of new and better products more efficient. The aim is that the test-rig also can be used as a tool for quality control between deliveries from the supplier. Throughout the project, previous research, users experience and the company's requirements and specification have served as a guideline for establishing a vision of what to develop. The test-rig has been constructed in terms of these requirements and specifications. Different load scenarios and sensors for measuring deflection and twisting of skis were tested throughout this project. This to observe the best possible solution adapted to pre-defined needs and specifications, as well as to increase my insight and learning. Test-rich concepts and load scenarios have been evaluated using evaluation matrices and conducted tests. The test-rig and sensor units are built as a result of this. Bending stiffness and torsional stiffness are calculated based on measurements made using these sensor units. Arduino has been used to program the various sensors used throughout the process. Python has been used to process data sampled by the sensor units. Tests have been conducted to measure bending- and torsional stiffness, and to verify the device's repeatability. The effect of mounted binding and ski-boot on the ski’s bending stiffness have been studied, based on the results obtained by F.Rainer and Kaps (2004), De Gobbi and Petrone (2008) and Federolf et al. (2006)

Denne oppgaven beskriver utviklingen av en modulær forskningsplattform, gjennom utviklingen av et fleksibelt eksperimentoppsett for forskning på simulatorer, brukergrensesnitt og fysiologi i menneskelige operatører. Bakgrunnen og motivasjonen for et slikt oppsett bygges hovedsakelig på argumentet om å tillate mer forskning i et høyere tempo. Vi sikter oss primært inn mot det gryende feltet affective engineering, der eksperimenter er en kritisk faktor og det er store kunnskapshull i dag. Gjennom å nyttiggjøre oss av fleksibilitet og gjenbruk, samt Wayfaring-metoden gjennom en fase med pilotstudier, søker vi å la eksperimenter bli gjennomført raskere og billigere enn i dag. Systemutviklingen er beskrevet i to faser. Først gjennom utviklingen av nødvendige moduler, både software og hardware, som behøves for i det overordnede systemet. Deretter beskrives en fase med pilotstudier, der systemtesting ble gjort gjennom å sette opp og utføre et demonstrasjons-eksperiment. Eksperimentet besto av et 2x2 ANOVA-oppsett og testet to ulike alarmsystemer under tilfeller av lav og høy arbeidsbelastning i en flysimulatorkontekst. Videre beskriver vi det endelige konseptet og systemkapabilitetene ved enden av pilotfasen. Spesiell oppmerksomhet ble da viet hvordan systemet oppnådde fleksibilitet og i hvilken grad systemfunksjonene kunne generaliseres ut av demonstrasjonskonteksten. Ved avslutning av utviklingsfasen lærte vi opp en ekstern forsker i bruk av systemet. Påfølgende dette fullførte hun selvstendig to gjennomkjøringer av demonstrasjonseksperimentet med suksess. Avslutningsvis i oppgaven diskuterer vi den utviklede plattformen og dens implikasjoner for videre forskning. En konferanseartikkel har blitt skrevet og innsendt basert på arbeidet i oppgaven. Artikkelen er gjengitt i masteroppgaven. This thesis describes the development of a modular research platform, through the development of a flexible experiment setup for simulator, user interface and human operators physiology research. The background and motivation for such a setup are built mainly on the argument of allowing more research to be performed quicker. We aim mainly at the emerging field of affective engineering, where experiments are a crucial factor and there are large knowledge gaps today. Leveraging flexibility and re-use as well as using the method of Wayfaring through a pilot study phase, we aim to allow experiments to be run quicker and cheaper than today. The system development is described in two phases. Firstly through the development of necessary modules, both physical and software, needed for the overall system to run. Secondly, through a pilot study phase where system testing was carried out by setting up and piloting a proof of concept experiment. The experiment was a 2x2 ANOVA setup testing two different alarm systems under high and low workload intensity in a flight simulator setting. Moreover, we describe the final concept and capabilities at the end of the pilot study phase. Special attention was devoted to how the system achieved flexibility and to what extent system functions were generalizable from the proof-of-concept context. At the conclusion of development, we trained an external researcher in using the system. Following this, she independently completed two runs of the proof-of-concept experiment with success. Lastly, we discuss the developed platform and its implications for further research. A conference article has been written and submitted based on the work in the thesis, and is reproduced within.

I samarbeid med St. Olavs og Biobank 1 har prosjektet til formål å prototype, iterere og teste et kuttekonsept og verktøy for kutting av prostataprøver som forberedelse til videre analyse (genuttrykksanalyse og MR-spektroskopi på vevsprøverfra prostata kjertler etter radikal prostatektomi). Basert på resultatene fra pre-masterstudien, er hoved teknologien som skal undersøkes, ultrasoniske frekvenser. Det er utført flere tester som har gitt kunnskap som er viktig for beslutningstaking under utviklingsprosessen. Aspekter som festemekanismer til skjæreblad, metoder for holding av prostata i en fast stilling under skjæreprosedyren, og måling av overføring av vibrasjoner, blir også vist og evaluert. Basert på litteraturvurdering og oppnådd kunnskap, blir potensielle farer ved bruk av høyfrekvent verktøy i denne sammenheng, evaluert. Testene utføres med en rekke testobjekter, men tester i menneskelig prostatavev forblir uprøvd In collaboration with St. Olavs and Biobank1 this project aims to prototype, iterate and test a cutting concept and tool for cutting prostate samples in preparation of further analysis (gene expressions analysis and MR spectroscopy on tissue samples from prostate glands after radical prostatectomy). Based on the results from the pre-master study, the core technology principle to be explored is ultrasonic cutting. Several tests have been performed that have provided knowledge essential for the decision making process during the development process. Aspects as fastening mechanisms of cutting blades, methods of locking the prostate in a fixed position during the cutting procedure and measuring transferring of vibrations, are also revealed and evaluated. Based on a literature review and acquired knowledge potential hazards are evaluated. Tests are performed with a variety of testing objects, but tests of the final prototype in human prostate tissue still remains.

Denne oppgaven er et arbeid som følger en prosjektoppgave som ble gjennomført høsten 2018. Den handler om samme tema. Denne oppgaven tar utgangspunkt i arbeidet gjort, og utfordrin- gen gikk av Brita Pukstad og hennes forskning på sårheling. Det er fortsatt i dag aspekter knyttet til mekanismene ved sårheling som ikke er fullt forstått. For å oppnå en dypere forståelse av sårhelingsprosessen, lager Pukstad hudmodeller in vitro og studerer hvordan disse modellene responderer på ulike miljøer. Denne in vitro hudmodellen er små 6 mm hudprøver som er tatt fra redundant hud fra pasienter som har gjennomført en hudreduksjon. Disse prøvene har 3 mm sår på seg. Disse er lagt i brønner med ulike løsninger for å bidra i sårhelingsprosessen etter at de er fjernet fra sin donor. For å øke gyldigheten til resultatene er det ønsket en standaris- ert metode for å lage denne hudmodellen. De eksisterende løsningene er enten gjort manuelt for hånd, eller de skaper ikke sår som er dype nok. Målet med denne oppgaven er utvikle et produkt som lager standardiserte sår for in vitro hudmodell. «Wayfaring» og. Hyppig prototyp- ingmetodikk er implementert for å oppnå dette. Oppgaven har fokusert på å teste en tidligere utviklet prototype kalt «wound generator 1.0» på menneskehud. Og basert på disse resultatene utviklingen av en ny, relativt høyoppløst proto- type for å lage standardiserte sår. Den Nye iterasjonen kalles «wound generator 2.0». Den er basert på et vakuumkammer som har som formål å virke som en motkraft mot huden som blir kuttet. Kuttingen blir gjort med en 3mm stansebiopsi festet til en roterende el-motor. Stansen og motoren er festet til en brakett som ved hjelp av to akslinger blir ledet opp og ned i z-retning. Z-forflytningen av braketten ar aktivert av to gjengede stenger som er aktivert av en steppermo- tor og som beveger braketten opp og ned og etterligner et ormegir. Når det nedgående kuttet er gjennomført vil et barberblad aktivert av en servomotor kutte horisontalt for å fjerne den 3 mm hudbiten og etterlate det standardiserte såret. Elektronikken kontrolleres av en Arduino som mater data om kuttedybde basert på data fra steppermotoren. Resultatet av oppgaven er mangelfull da det ikke har vært anledning til å teste versjon 2.0 på menneskehud. Testen ble gjennomført på kyllinglår. Egenskapene til kyllinghus og menneske- hud er veldig forskjellige. Dette er vist i den tidligere prosjektoppgaven. Resultatene er derfor mangelfulle. Informasjon og resultater fra tester gjort på versjon 1.0 kan overføres til testingen gjort av 2.0. The kan konkluderes med at deler av det mekaniske designet av 2.0 fungerer godt, og at det på mange områder er en vesentlig oppgradering fra 1.0. Resultatene tyder også på at man med noen justeringer vil ha en lovende løsning også til menneskehud. Det er også funn som tyder på at noen endringen bør gjøres. This thesis is the work following a project thesis conducted on the same subject in the fall of 2018. This thesis is taking basis in the work done, and challenge proposed, by Brita Pukstad and her research on wound healing. It is still to this day aspects of the mechanisms in wound healing that is not fully understood. In order to gain a better understanding of the wound healing pro- cess, Pukstad is creating In Vitro skin models and studying how these models behave in different conditions. The In vitro skin models are small 6 mm diameter skin samples taken from redun- dant skin of skin reduction patients. The skin samples have 3 mm wounds inflicted on them. These are laid in wells with different solutions to aid the healing process of the skin model after it is separated from its donor. To improve the viability of the results it is required a standardized method of creating the skin model. The existing solutions are either done manually by hand, and would not qualify as sufficiently standardized, or the wounds is not penetrating sufficiently deep into the skin. The goal of this thesis is to develop a solutions for creating standardized wounds for In vitro skin models. Wayfaring (design thinking) and rapid prototyping methodol- ogy is implemented to achieve this. The thesis have been centered around testing a previously developed prototype the Wound Gen- erator 1.0 on human skin. And based on these results developing a relatively high resolution prototype to create standardized wounds. The device new iteration is the Wound Generator 2.0. It is based on a vacuum chamber that has the purpose of by a counter force acting on the skin being cut. The cutting is being done by a 3 mm punch biopsy blade mounted to a rotating DC motor. Blade and motor is fastened to a bracket that is guided by two guide rails to move in the z-direction. The z-displacement of the bracket is actuated by two threaded rods that in turn is actuated by a stepper motor. The rotation of the stepper motor moves the bracket up and down, acting as a worm gear. When the downward cut is completed a razor blade actuated by a servo motor is cutting horizontally to remove the 3mm skin and leaving the standardized wound. The electronics are controlled by an Arduino which feeds data on cutting depth, based on the step- per’s travel, to a computer. The results of the thesis is lacking due to not being able to perform final testing of the 2.0 on human skin. The test were performed on chicken thigh with skin. The characteristics of chicken skin and flesh is very different of that of the human skin as was shown in the prior work in the project thesis. The results are therefore lacking. Information and results from testing 1.0 can however be deducted into the testing of 2.0. It can be concluded that parts of the mechanical design of 2.0 is functioning properly, and an substantial improvement in many regards to 1.0. The results also suggest that it with some adjustments would be a viable solution also on human skin. There are also finds suggesting that some changes of the design should be implemented.

In product development and engineering design research, discoveries have revealed the benefits of using prototypes and prototyping to increase the probability of successfully making new innovative and novel products. With the increasing competition in the global markets, innovation is key for sustainability and competitiveness. In the early stage of product development, large solution spaces are explored along with ambiguous information and uncertainty. This critical stage will greatly affect the cost and quality of later development activities, such as optimization and manufacturing. This thesis aims to augment physical prototype activities in the early stages of product development by introducing state-of-the-art technologies and alternative design approaches, to make the process simpler and faster in many scenarios. The goal is to allow the generation of more design iterations in the pre-requirement stages, ultimately increasing the probability of making better products faster. The main methods explored are photogrammetry-based 3D scanning, converting hand drawn sketches to physical parts, documenting microcontroller-based prototypes and using object recognition frameworks. These methods have been developed and implemented into a physical system called Protobooth v2, that users can interact with to make prototypes and accelerate learning, intended for both experienced and inexperienced product developers and teams. This system builds on an ongoing research project at the TrollLabs laboratory at NTNU in Trondheim, aiming to capture prototypes to enable research on early stage product development. Each of the main methods have been developed into functional prototypes and are presented with experiments and practical examples. They have been implemented with the physical Protobooth v2 system. Using Protobooth v2 is simple and it demonstrates a potential for augmenting physical prototype activities, by converting complex models to digital representations and documenting microcontroller output. Additionally, by utilizing computer vision and artificial neural networks, the system can automatically detect and recognize components to provide useful information about the components while building prototypes.

This thesis aims to provide an understanding of the Why? and How? of a modern toolmaker, specifically within the biomedical sector, on a theoret- ical and practical level and what their role is within a product development process. Over the four years of this PhD work, a total of ten projects were completed by - or with the involvement of - the author. Framed by the grounded theory, these project provide the fundamental data for the conclusions of this thesis. In addition, a total of thirteen scientific contri- butions were - and will be - published that cover the three cornerstones of the toolmaker: The role of users, the importance of prototyping and how to successfully do it, and how this is used in the biomedical sector. The structure of this thesis is closely following the three essential stages of product development process as encountered when using the Wayfaring method: Orienting, Probing, and Bringing it Home. In addition to high- lighting key-insights from the projects to emphasize the importance of each stage, the thesis is following the development path of the main project, a novel flow-chamber setup for cell research that was developed by the author. This project allows to reflect on a complete product development process from paper prototype to patent-pending beta-version, with a focus on why the toolmaker made all the difference. The conclusions of this thesis are, given the nature of the scientific method, of qualitative nature and address the initially stated questions of Why? the toolmaker had an impact in the projects and How? one can implement an according role within a product development team.

Periodontal disease is the leading cause of tooth loss in adults. Diagnosis of periodontal disease varies from visual inspection to invasive probing of the gums. In this master thesis the goal is to enable the use of ultrasound imaging to assist in diagnosis and classification of periodontal disease. Utilizing ultrasound could allow for a non-invasive, simple andp ainless imaging technique. Through dialog with dental and ultrasound experts the latest equipment and knowledge was sourced. As there has not been any widespread use of ultrasound in dental imaging, a novel sleeve system for in vivo imaging has been developed to further investigate the potential of ultrasound. Utilizing the sleeve resulted in improvedimage quality that could not be matched by modern post-processing algorithms. The dentalstructure can be clearly seen, but to further investigate the possibilities a clinical trial isrequired to acquire images of unhealthy periodontium. The main objective of this thesis isto advance the knowledge about ultrasound in dental imaging, and be as a stepping stonebetween the initial idea and the start of a research project at NTNU.

The correlation between ski friction and a ski base surface that can enhance hydrophobicity by geometry has been investigated. A review is given of existing knowledge about the linkage between ski friction mechanisms, generation of a water film between ski-and snow base, and wettability. The effect of roughness induced wettability is explained, and forms the basis of examining nanostructuring techniques to obtain a hydrophobic ski base. The fabrication of an anodic aluminum oxide (AAO) miniature ski prototype is reported, conducted with a simple and low-cost setup. Anodization is performed in a single step procedure with a constant current of 5.18 A for 40 minutes, and an electrolyte consisting of sulfuric acid, oxalic acid, and sodium chloride. The resulting porous surface structure was analyzed with SEM, and pore diameters measured to 20-33 nm. The contact angle (CA) was measured to 130° for the anodized sample when coated with PTFE. A reference sample with identical dimensions, but with a longitudinally milled surface, was fabricated and measured a CA of 92°, also when PTFE coated. Friction tests were performed with a Linear Analyzer of Road Surface (LARS) apparatus in which the samples were accelerated on a linear snow track at an indoor facility. The coefficient of friction (COF) was measured at a constant velocity of 2 m/s. Measurements were conducted on both anodized and reference sample at snow temperatures of -10 °C and -4.6 °C. The surfaces were analyzed with an optical profilometer (OP) and CA measured, both before and after friction tests. The results from the friction tests state that the anodized sample holds lowest COF at -10 °C, but the reference sample has lowest values at -4.6 °C. However, it is believed that the observed wear of PTFE is more critical for the AAO sample, and hence a clear conclusion cannot be stated on whether the nanoporous hydrophobic surface yields reduced friction compared to the reference sample.

Key insights from the making of two stators, and two complete electric motors, have laid the foundations for an improved production machinery based on the patent of Alva Motor Solutions. As resources are limited, an iterative product development process driven by probing and wayfaring was applied to meet the pressing milestones faced by the startup company. Through this research Alva has honed the ability to construct competitive electrical motors with semi-automated production equipment. The course towards industrialization is set, and further work suggested.

By cutting specific patterns in thin sheets of material it is possible to acquire a high degree of compliance in and out of the plane. These structures, called flexure patterns, are often applied in situations causing large deformations and contact. This is known from theory to introduce a non-linear behavior, in which the load and deformation no longer has a linear relation. This thesis sets out to analyze these sources of non-linearity, and find their significance in design applications. This is done by performing a parametric study of the base repeating unit of two known patterns: the LET/Slits pattern and the YdX. The goal is to contribute to the knowledge and understanding of the mechanical behaviour of flexure patterns. Non-linear simulations with the FE-software ABAQUS have been used in combination with python to make the analyses as automatic as possible. The results show that for some load cases there are small deviation between the linear and the non-linear simulation. The deformation mechanisms of important structural members are assumed to play significant roles as sources of geometric non-linearities. In addition, the results indicate that an automatic analysis of contact non-linearities would require further work and is a complete study by its own.

Paper 1 encompass a literature investigation and could aid in understanding the business model phenomena on an overarching process level and how these processes could be improved. Paper 2 encompass an in-depth case study and could aid in understanding the business model phenomena on an organizational and network level. Paper 3 encompass statistical testing of company data and could aid in understanding the business model phenomena on a company level and in-between companies. Paper 4 encompass a user experiment and could aid in understanding the business model phenomena at a product level. Conducting multiple methods poses great requirement on a researcher. It may be next to impossible for one single researcher to conduct the methods properly. Academic rigor should not be compromised. Therefore, the handover between the methodological approaches and the specific methods must be appropriate. Complex, uninvestigated phenomena and context require a genuine mixed method approach, which research designs with a solitary purpose fall short on. A deliberate selection of methods combined with applying the flexibility principle to research design results in a research design that is not preset, but that adapts to the research process as it progresses. A flexible research design can further enhance the mixed method research approach s ability to investigate broad and complex phenomena such as the context in this thesis.

A python package for robot communication and a new predictive display based on image transformation has been developed. An experiment was performed and this thesis present the results.

The aim of this thesis Development of the Market s Lightest Backcountry Ski Binding is to see how topology optimization can be used to create a light weight and stiff ski binding. Sports equipment is getting lighter and lighter, making it increasingly important to use material as effectively as possible. Topology optimization is not a new technology, but the rapid development of constraint free production methods, like additive manufacturing, creates new possibilities which now makes topology optimization more relevant in mechanical engineering than ever. This thesis firstly examines the mountain ski binding market, user scenarios and relevant design variables. In the second stage, a set of optimizations are done to gain knowledge and experience, to define the test setup and to ensure that the optimized results will be as realistic as possible. Further, a set of models with different combinations of design variables (boundary conditions and design space dimensions) are topology optimized. The results are tested for both static and cyclic loading and redesigned to reduce von Mises stresses and increase the fatigue life. Prototypes are then manufactured and tested to verify the design. The result is a binding with a weight of 26.70 grams, which is lower than the current market leader. Including the design variables in the optimization halved the strain energy, thus making the binding stiffer, from the initial topology optimizations. Several topology optimization cases are studied and in many of these a fixed setup is topology optimized, leaving boundary conditions and design space dimensions unchanged in the analysis although there is a huge potential in optimizing these as well. These discoveries result in the categorization of three different approaches and the identification of several limitations which are discussed in the paper State of the Art of Generative Design and Topology Optimization and Potential Research Needs (Tyflopoulos, Flem, Steinert, & Olsen, 2018). This paper was accepted at the NORDESIGN 2018 conference.

In search for more efficient propulsion through water, many methods for improving on the existing hydrofoil and propeller designs we use today have been suggested, the topics of adaptive hydrofoils and propulsion systems is increasingly popular. A large portion of all world trade happens by sea, and even incremental improvements in propulsive efficiencies can have big economic and environmental impacts. Many researchers have looked to nature to find answers as to how to more efficiently use and manipulate water for propulsion and maneuvering. In this master s thesis I investigate the mechanisms that enables fish to use flows actively, and efficiently, to improve the capabilities of adaptive hydrofoils through sensory inputs. The thesis is based on the work done in my pre-master s thesis into the effects of Kármán gaiting, where fish use alternating vortices shed from obstructions in flows to generate forward momentum and is a response to the limited capabilities I found in the field of high density, non-intrusive/flow-obstructive, flow sensing at the surface of bodies and hydrofoils, such that correct adaptive movements can be made. I investigate different ways to sense flows along the skin of hydrofoils suspended in flowing water through prototyping and comparing solutions. Through comparison the piezo resistive effect of carbon fibers mixed with rubber silicone showed superior capabilities in detecting pressure changes as well as its ability to be easily imbedded and cast into flexible hydrofoils. The composite was further investigated in its capabilities in flow detection, through expanding the amount of datapoints with multiplexers to generate bigger pictures of the flow and enabling tendencies and pressure gradients to be seen. A water tunnel was run at different flow velocities of 0.04m/s and 0.1m/s, both free flowing and altered to generate alternating Kármán vortex streets. Two silicone hydrofoils were fitted with 64 sensors each. At 0.1m/s the Kármán flow generated cyclic behaviors in the forward-facing sensors in one of the hydrofoils, that directly corresponded with the observed and calculated shedding frequency. A dynamic neural network was further able to distinguish differences between all flow scenarios in separate data from the same calibrations and setups that were used in the training sets. The carbon fiber rubber silicone sensor shows similar piezo resistive properties as the popular carbon nanocomposite-silicone sensors, but the ease of manufacturing and low cost of the carbon fiber sensor could open up new possibilities and implementation in areas where such carbon nanocomposite sensors have not yet been deployed.

This master's thesis describes the development of a hydroponic home growing system. The system is meant be utilized by both users with and without any pre-existing experience in hydroponic growth. To achieve this, a user-centered design that is easy to operate and assemble, have been one of the main goals during the development process. Relevant technologies and existing research have been identified in order to understand the inherent features and components necessary in order to develop a hydroponic system. The project has utilized wayfaring as its governing product development methodology, and concept generation, prototyping and testing, have guided the development process forward. This all lead to a proposed system and a prototype that proves the validity of its concepts. The author was engaged in this project from January 2018 until the thesis was delivered in July 2018.

A flexure pattern increases the ability of a plate to undergo large elastic deformations. It is characterized as a 2D mechanical metamaterial that consist of flexures configured in a pattern that increase the compliance compared to the bulk material of which it has been made. This work contributes to the basic understanding of flexure patterns by describing the geometry, symmetries, principle deformation mechanisms and anisotropic elastic properties, using of the computational homogenization technique. The criteria for diagonal, orthogonal, tetragonal and isogonal elasticity is categorized though the minimal symmetries found in the pattern. Various examples of auxetic (materials with negative Poisson's ratio), orthogonal and isogonal patterns are explored and the relations between geometrical parameters and elastic properties are found. A set of design principles and methods are compiled to a framework that enables creation of new patterns from a set of desired behaviors.

Rethinking Surface Design and Materials of Skis

The world is becoming increasingly sedentary. Many jobs centered around physical work are being automated, and the time spent physically active is sinking. Various CPM (Continuous Passive Motion) chairs, and other chairs designed for increasing the activity-while-seated level have been made, and research on them are somewhat conflicted, with a majority not finding a statistically significant difference in the in the level of activity while seated (L. L. van Deursen et al. 1999; O Sullivan, O Sullivan, O Keeffe, et al. 2013; O Sullivan, O Keeffe, et al. 2012; Aota et al. 2007; Groenesteijn et al. 2012). Building on the previous work done the project thesis (appendix D) , and in TMM4245 The Fuzzy Front end 2016, two prototypes have been built capable of classifying sitting postures, estimating activity levels and altering tilt resistance. This thesis documents the building of the two prototype chairs, designed to test the effect of automated tilt resistance on a user s level of activity while seated. One chair is built to be immediate, and the tilt resistance alter the neutral angle of the chair, the other with a subtler change. A small pilot test was performed ensuring that the prototypes were working as intended, but the limited testing did not yield enough data to determine the effects either prototype could have on seating behavior. Further field tests are strongly recommended. It is also recommended to implement a reinforcement learning algorithm to the tilt resistance setting. The result could be very interesting both commercially, and with regards to research on human computer interaction.

This thesis goes into early-stage (Fuzzy Front End) concept generation for comprehensive, multidisciplinary projects in extreme environments, such as a fully autonomous unit performing anode covering in an aluminum electrolysis plant. Through cycles of low resolution prototyping of product concepts and relevant environment effects, the author elaborates assumed objectives related to the operation and the hazardous pot environment. The development process applies high flexibility, with rapid, qualitative testing of prototypes. This may continuously change or dispose initial objectives for the product, as a natural part of the concept detailing. This guides the suggested product and operation concepts towards accomplishing objectives with sufficiency.

This master's thesis describes the development of a conceptual prototype of a smooth actuation and control system of the eyes in a healthcare simulation mannequin with the aim to achieve perceived human like eye movement and behavior. The chapters of this thesis identifies, collects, judges and conceptualizes technologies to simulate the eye movement and behavior. The project identifies and develops a system that achieves realistic human-machine interactions, without a feeling of eeriness or unease, as is common with humanoid simulations. Prototypes are shown that tackle the problems of human-like movement and behavior. The system prototype allows the operator to look through the eyes of the mannequin by displaying the mannequin's view to the operator in the head mounted device. The head mounted device processes and captures the eye movement of the operator and mimics this in mechanical actuation of the eyes of the mannequin. By letting the operator see the environment around the mannequin, the operator is able to interact with medical personnel through the mannequin. Possibilities for automation of the behavior has been proposed, where the most promising is supervised learning in artificial neural networks. By using deep learning, it may be possible to train an algorithm to mimic human behavior and thus removing the need for operator control. The system that has been proposed is an alpha prototype that has implemented critical functions for operation and proof of concept. Implementation is dependent on further work and optimization. The system has the potential to provide further dimensions of simulations by introducing nonverbal communication and diagnostic tools that use the eyes.

This research is part of a study of autonomous tugboats. The study is a cooperation between two students, but the projects are separate. The common goal for the study is to develop a working prototype of an autonomous tugboat swarm. In this part of the study, algorithms for planning the shortest paths for the tugboats have been developed. A literature review has been conducted to find the best algorithms and framework for the path planning system. The environment has been simplified by using a road map approach called ''visibility graph''. The shortest path within this graph is found by using A* search algorithm. This will give the shortest path within the non-simplified environment, if one exist. Algorithms from external libraries have been adjusted to fit the tugboat framework. Some additional algorithms have also been developed, for instance a new method to find paths around a ship in real-time, only knowing its dimensions, position and orientation. In addition, a speed-up and memory saving version of the shortest path algorithm has been developed, by calculating all shortest paths to and from all points offline. The speed-up was measured in five different test environments, and was measured to be on average 12.3 times faster than ordinary A* search in a visibility graph. The speed-up becomes even more impressive as the complexity of the environment increases. The path planning system and communication has been implemented using the ''Robot Operating System''. This system has been tested successfully on a physical test setup. The physical test setup is a starting-point for future research of autonomous tugboats. With this setup, the development in the study can be conducted using a bottom-up approach. Rather than starting off by creating a system seeming to perfect in theory, this setup can be used for rapid testing of smaller features.

Flapping Foil is an ongoing development project at NTNU Technology Transfer working on a new way of propulsion for boats. Their goal is to eliminate two of the biggest issues with boats today, the low efficiency of boat propellers and the high friction created when moving in the water surface. Their solution to these problems is a hydrofoil propulsion system that will in addition to propel the boat more efficiently also lift it up out of the water reducing friction. This master thesis has in collaboration with Flapping Foil looked at the possibility of using a curving hydrofoil for the propulsion system instead of a more traditional hydrofoil with a flap attached to it's rear end. There have been two objectives of this, to develop a prototype for this kind of hydrofoil and to create a test setup for comparing the prototype to a traditional hydrofoil. Unfortunately because of a failure in an electrical component vital to the test rig no data has been collected, however all planning is completed and testing is ready to be done should the component be replaced. In addition to this a working prototype for a curving hydrofoil has been constructed.

Plastic is a very broad family of materials that may provide a wide array of mechanical properties depending on the plastic or production method in question. This is why many industries have chosen plastic as their material of choice for the production of anything from plastic bags to underground piping. There is, however, a prominent issue concerning the heavy environmental impact of plastic. This is both due to the processing of crude oil and lack of biodegradability which in turn impact nature if the plastic is not processed in a proper waste disposal stream. In some applications such as food trays and food containers, moulded pulp has been used as a sustainable and more environmentally friendly alternative to conventional plastic products for quite some time already. The applications of moulded pulp do however not expand very far at the current time and that is why this thesis seeks to assess the properties of moulded pulp samples to evaluate the potential for replacing plastics in new mass production settings. Moulded pulp does not only have an environmental motivation but it also shows promise of being more economically viable in certain settings and the paper industry is always looking for new applications for them to use the pulp they are manufacturing. There are, of course, many applications where it is not viable or desired to replace plastic at all but the motivation is strong for replacing plastics in applications where it is possible. This is especially due to a general shift in society towards environmental awareness which is a major driving force in changing industries. If people want change, the industry will in most cases have to adapt in order to remain profitable. For this thesis, the focus was on assessing the mechanical properties of moulded pulp and how the fibres and the material behave when thermoformed in a fashion similar to that of thermoformed plastic. This was primarily done through the means of tensile testing of paper sheets and moulded specimens. In addition to this, the effect of reinforcement from microfibrillated cellulose (MFC) was also explored. In the end, the results of all the tests were very promising and the strength of the moulded samples was on par with quite a few of the most common plastics used today. The thermoforming, which is one of the most promising methods for moulding paper, also showed its capabilities of producing excellent samples and the future development of thermomoulding processes could result in highly efficient moulding methods down the line. This thesis did not have any specific applications in mind when the research was conducted but when considering the results it is clear that with further research, this material could possibly be used for a wide array of applications and it could possibly be used in a laminate construction to greatly enhance the strength and make the material viable for even structural components.

Human-Computer Interaction systems are rapidly changing. In such interactions, human performance might be essential for the system to function in the best possible way. Thus the need to optimize for interaction design that take human emotion in to account are advancing. Great challenges are involved in capturing emotions, for instance that human emotion are complex, and difficult to quantify. In this thesis theory about body language and emotions are adapted in an attempt to introduce a new tool for Affective Engineering. This is done through developing a pilot chair sensor setup and test its capabilities towards capturing emotional states. A pilot experiment setup is used to test association between posture sensed by the chair and level of mental activation or arousal. Statistical analyzes on the data recorded is done to compare established measures of arousal with posture data from the chair. The analyzes show some tendencies toward association between movement and arousal, but more comprehensive analyzes are needed to introduce the setup as a new tool for affective research. The main take away from this thesis would be the experimental framework that was developed, which with small modification can be used for testing tools to be used in interaction studies. Keywords: Affective Engineering, Arousal, Body Language, Emotions, Interaction Studies, Unembedded Sensing