Search

Sustaining real-world human-robot interactions requires robots to be sensitive to human behavioural idiosyncrasies and adapt their perception and behaviour models to cater to these individual preferences. For affective robots, this entails learning to adapt to individual affective behaviour to offer a personalised interaction experience to each individual. Continual Learning (CL) has been shown to enable real-time adaptation in agents, allowing them to learn with incrementally acquired data while preserving past knowledge. In this work, we present a novel framework for real-world application of CL for modelling personalised human-robot interactions using a CL-based affect perception mechanism. To evaluate the proposed framework, we undertake a proof-of-concept user study with 20 participants interacting with the Pepper robot using three variants of interaction behaviour: static and scripted, using affect-based adaptation without personalisation, and using affect-based adaptation with continual personalisation. Our results demonstrate a clear preference in the participants for CL-based continual personalisation with significant improvements observed in the robot's anthropomorphism, animacy and likeability ratings as well as the interactions being rated significantly higher for warmth and comfort as the robot is rated as significantly better at understanding how the participants feel., Comment: Accepted at the 10th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW). 12 pages, 7 figures

Recent research is emerging in the field of Social Robotics where robots have the potential to serve as tools to improve human well-being. However, research exploring the expectations and perceptions of prospective users of such robots, and the professionals who currently deliver these interventions, is limited. In this paper, we present qualitative analysis of discussions with prospective users and experienced coaches regarding the design of robot well-being coaches. We invited participants interested in well-being practices to take-part in a Participatory Design (PD) study, consisting of individual interviews and a focus group discussion (N P = 8). Discussions focused on ideating how a robot could function as a mental well-being coach, based on their experiences with well-being practices. Data triangulation was employed by interviewing three professional coaches as additional sources of information. This resulted in a rich set of data, which we transcribed and analysed using Thematic Analysis (TA). The developed themes regarding robot features, form, behaviours, robot-led well-being practices, and the advantages and disadvantages these could provide, were compiled and are discussed in detail. We present this data together with tabulated quotes from the participants and coaches, to pave the way towards designing robot coaches that can provide supportive interventions to improve the mental health and well-being of their users.

Most industries are shifting from product-oriented business models towards services to step up the value chain and engage in long-term relationships with their customers throughout the service lifecycle. Digital technologies are contributing to servitization in many ways by creating and enabling capabilities like connectivity, IoT, data generation and assessment, etc., for new value generation, distribution, and capture. Because value is subjective, dynamic, and changes during the service lifecycle, service providers need to examine closely the value perceptions of customers to constantly provide better value and remain relevant with the competition. Through a consumer survey and a longitudinal study of thirteen customers, this paper uses qualitative and quantitative assessment to identify the value dimensions that play a major role for customers being onboarded on a digital enabled service, and also highlights how customer value dimensions change over the course of the service lifecycle. One important finding is that change in customer value perception does not follow a pattern and is highly individual and personal. This opens a discussion regarding the need for hyper-personalization in successful servitization, and the role of digital technologies towards the same.

Cyclic transport missions involving fleets of vehicles are common in quarry and mining operations and have shown to have a significant potential for energy optimization. Vehicles such as articulated haulers and dump trucks utilized in transport missions can be of different brand, type and have different performance characteristics such as engine power, traction and load. Vehicles may also be operated and owned by different organizations as they can be subcontracted to an operation. The transport operation characteristics include stochastic behaviors and activity times that fluctuate over time, and hence real-time control is required for efficient optimization. As the vehicles are mobile, wireless communication also needs to be available. As the constituent systems (vehicles) have managerial and operational independence, a system of systems approach is applicable. This paper provides an overview of the key characteristics and requirements for such a system and discusses the pros and cons of acknowledged and directed system architectures. Further, a case study is presented where an acknowledged system of systems is implemented in a real world mine and evaluated through a qualitative assessment of an operator assistive optimization system. Key findings include the drawbacks and characteristics of the architecture approaches. Further challenges and potential for automation and production control of a larger process system are described.

A collaborative system of systems (SoS) is formed when independent organizations decide to cooperate to achieve mutual benefits, while retaining independence of their respective systems. Each constituent system (CS) of the SoS has a set of capabilities, some of which they agree to potentially use in active collaboration with others. Such an active collaboration is called a constellation and can be seen as an instantiation of the SoS which is created to provide a joint capability. Constellations are thus the working-horses of the SoS, but due to the operational independence of the CS, they have a choice whether to join a certain constellation or not. This paper discusses the reasoning and world model that is necessary for a CS to make well-informed decisions to join and leave constellations. We argue that it is necessary for the CS to understand not only the surrounding environment, but also to have models of other CS’ world models as well as of their probable future actions. It must be possible to predict whether other participants will uphold their parts of the collaboration or may defect from it to join another more rewarding constellation despite the agreements made when joining the SoS. The reasoning in the paper is illustrated using examples from two different collaborative SoS in the transportation domain.

Systems of systems arise when independently owned, operated and developed systems can achieve mutual benefits by working together. In collaborative systems of systems, there is no directing entity that instructs others how and when to collaborate. Instead, the collaboration relies on independent decisions by the constituent systems to form collaborating constellations, and the benefits are emergent properties of this. In this paper, we describe a design method for engineering collaborative systems of systems. We apply the method to the design of a collaborative system of systems for mobility in a rural setting close to an urban area and to truck platooning.

Small Unmanned Aerial Vehicle (UAV) platforms equipped with compact laser scanners provides a low-cost option for many applications, including surveillance, mapping, and reconnaissance. For these applications, semantic segmentation or semantic labeling of each point in the lidar point cloud, is important for scene-understanding. In this work, we evaluate methods for semantic segmentation of three-dimensional (3D) point clouds of outdoor scenes measured with a laser scanner mounted on a small UAV. We compare the performance of four different semantic segmentation methods, which are all applied in a scan-by-scan fashion, on semi-sparse laser data. The best method achieves 95.3% on the three classes ground, vegetation, and vehicle in terms of mean intersection over union (mIoU) on a previously unseen scene from a different geographical area. The results demonstrate that it is possible to achieve good performance on the semantic segmentation task on data measured using a combination of a small UAV and a compact laser scanner.

The boom of Internet of Things (IoT) devices has brought along new security concerns which earlier were not thought of. This has expanded the potential for digital forensic investigators to gather rich evidence from these sources. The data on these IoT devices is not easily accessible due to the lack of proper techniques to investigate such devices. This paper presents a sophisticated non-invasive method to investigate IoT devices. The software activities running on a device can be inspected by observing the electromagnetic radiation emitted from the devices during the process. The objective of this project is to evaluate if it is possible to classify the software activities being run on an IoT device by performing an electromagnetic side-channel analysis (EM-SCA). This paper presents a methodology for analyzing the EM side-channels and classifying encryption algorithms being run on a Raspberry Pi 4. This work demonstrates that the cryptographic encryptions can be classified with over 95% accuracy by using deep neural network-based classifiers.

The constituent systems (CS) that together form a system of systems (SoS) have a continuous need to assess situations and make decisions. In addition to operating in the environment, they also need to decide upon their status in the system of systems and be aware of their relations to constellations in the system of systems. To be able to make the best possible decisions, the constituent systems need to have an accurate situation awareness, i. e., an understanding of the environment they are in, what other elements are present therein and how this will develop in the future. In this paper we present an analysis of the situation awareness needs of a constituent system in different stages of its life-cycle: ignorant of the SoS; prepared forjoining the SoS; a passive CS; an active CS. A conclusion is that CS need to have world models that include information about other objects. Examples from the truck platooning application are given.

Online education is changing how teaching is done at universities and provides new opportunities to reach out to practitioners. In this paper, the development of an online course in systems-of-systems engineering is presented, as well as results from the first instance of the course. The paper describes how the course was designed; how it was produced; and experiences from giving it. Challenges with online education in the systems engineering subjects are also highlighted.

This paper puts forward the hypothesis that all systems-of-systems (SoS) need to deal with geospatial information. It discusses some fundamental aspects of such geodata, including entities, coordinate systems, features, and representation. It then presents how geodata can be used for various purposes in SoS and suggests architectural strategies for handling geodata in this context, including the use of linked data to represent both geodata and other information; triple stores for databases; and cloud servers for executing geodata related constituent system functionality.

We present a system for forest mapping from an autonomous UAV, flying below the tree canopy. The UAV is equipped with four stereo cameras, which provide information for navigation, mapping, and collision avoidance. The main focus of the paper is positioning and forest mapping, but the full system, including both hardware and algorithms, is described.

We present the participatory design process of a robotic tutor of assistive sign language for children with autism spectrum disorder (ASD). Robots have been used in autism therapy, and to teach sign language to neurotypical children. The application of teaching assistive sign language - the most common form of assistive and augmentative communication used by people with ASD - is novel. The robot's function is to prompt children to imitate the assistive signs that it performs. The robot was therefore co-designed to appeal to children with ASD, taking into account the characteristics of ASD during the design process: impaired language and communication, impaired social behavior, and narrow flexibility in daily activities. To accommodate these characteristics, a multidisciplinary team defined design guidelines specific to robots for children with ASD, which were followed in the participatory design process. With a pilot study where the robot prompted children to imitate nine assistive signs, we found support forthe effectiveness of the design. The children successfully imitated the robot and kept their focus on it, as measured by their eye gaze. Children and their companions reported positive experiences with the robot, and companions evaluated it as potentially useful, suggesting that robotic devices could be used to teach assistive sign language to children with ASD.

In the field of systems-of-systems (SoS) engineering, there is broad agreement on a few characterizations, and these are currently being standardized. However, many aspects in the field still lack an established terminology. In particular, there are unclarities related to the internal structure of an SoS, and on the internal states of constituent systems. In this paper, a refined terminology is therefore proposed, which covers the internal substructure of an SoS; the states of constituent systems in relation to those substructures; and how it relates to hierarchical levels. This terminology can also be used to classify the characteristics of an SoS through different metrics. The terminology is illustrated through three examples of SoS applications in various domains.

This paper addresses the problem of exchanging complex data between the constituent systems in a system-of-systems. This is necessary to ensure that they have compatible understandings of the world surrounding them and entails a need for semantic interoperability between the constituents. Through a case study of a road construction system-of-systems, the world wide web technologies of linked data and ontologies are explored as a framework for data representation and exchange. This data includes several broad categories, such as assets, interfaces, organizations, capabilities, missions, and observations, as well as various properties of those. It is also discussed how the constituents can use this data for reasoning and decision making. The results have been validated through a simulation-based research prototype of the road construction case, from which experiences are reported.

Hyperspectral imaging based on a linear variable filter (LVF) close to the focal plane array (FPA) is an attractive technique since it gives a small, lightweight, and inexpensive sensor. Linear variable techniques can capture the full spectrum within a given spectral range in one image at the same time, but only records one wavelength in each pixel. Reconstruction of a spectrum for a point in the scene requires fusion of spectral information from a large number of registered images. We present a procedure to reconstruct a hyperspectral data cube by registering a large number of overlapping images with spectrally varying content using feature-based registration and transforming them to a joint coordinate system. The accuracy of the registration is evaluated and shown to yield a reconstruction error of less than one pixel.

Road construction is a very large business segment, consuming enormous public funding every year and with significant environmental impacts. However, the rate of efficiency improvement during the last few decades has been negligible, whereas other industries, such as manufacturing, have seen very large improvements by applying automation and Leanbased flow optimization across the production system. In this paper, we outline a system-of-systems concept for road construction which applies similar principles as have previously proved successful in other industries. The paper identifies efficiency attributes and wastes in current practices, which lead to a conceptual solution that focuses on improved coordination of working machines. Technical elements from Industry 4.0 are considered as potential building blocks in this concept, identifying similarities and differences between the construction domain and other industries. Finally, challenges are identified, in particular within knowledge representation and information management.

Urbanization is one of the major current trends in society. Cities around the world are looking into “smart” solutions based on information and communication technology to deal with the challenges that result from this development. Mobility is one of the most important areas to address, and system-of-systems solutions where vehicles and infrastructure are connected have a potential to improve urban transportation in many aspects. In this paper, current initiatives related to mobility in smart cities around the world are surveyed, and this is complemented with input from focus groups of transportation stakeholders to identify the important aspects of the problem. Based on this, challenges related to the application of systems-ofsystems in urban mobility are identified.

As systems-of-systems start to be more common in commercial applications, an analysis of emergent properties related to utility and cost for all stakeholders becomes critical. This paper describes an approach for this based on network simulation and game theory, which is applied to truck highway platooning. The emergent properties of energy efficiency and transport efficiency are studied as a consequence of the strategies for platoon formation, and it is shown that having information about the route planning of other vehicles has a significant positive effect on the properties. This indicates the need for a mediator in the platooning system-of-systems architecture. Its role is to communicate plans of the constituent systems.

The characteristics of systems-of-systems (SoS) present fundamental challenges regarding properties such as safety, security, reliability, and robustness. This is due to the SoS nature where a collection of independent systems cooperate to fulfil certain high-level objectives. Risk analysis is thus an important activity in SoS engineering. This paper presents a risk analysis method which extends the existing STAMP safety analysis method that is based on systems thinking. Our extensions are aimed at coping with other risks than safety, and the usage is tailored to SoS. The method aims at deriving requirements on the constituent systems that will reduce the emergent risks on the SoS as a whole. The method has been applied to a case study of vehicle platooning.

Three-dimensional (3D) reconstruction of buildings is an active research area with applications in e.g. city planning, environmental simulations, and city navigation. Automatic 3D building reconstruction methods based on point clouds from laser scanning or methods based on high resolution dense photogrammetric point clouds are common in the literature. In applications where large land areas need to be covered regularly it is not practical to use laser scanning or acquire images with high resolution and large image overlaps. In these applications the reconstructed photogrammetric point cloud has low resolution with less building details. We present a method where the most common roof types are classified using a deep convolutional neutral network (CNN) pre-trained using RGB data in this challenging type of data. In addition’ a method for roof height estimation for each roof type is presented to support automatic 3D building reconstruction using model building shapes. Results are shown for a low resolution dense photogrammetric point cloud generated using multi-view stereo reconstruction of standard overlapping aerial images from nationwide data collection. The method is intended to support automated generation of a nationwide 3D landscape model.

Bin packing problem has gained a wide interest in academia and in practice since the problems quickly becomes intractable as the problem grows. In this paper two models are compared: one model that generates exact solutions for bin packing problem and one that uses the same fundamental approach on the problem but extended with a heuristic combination of next-fit and a combinational best-fit. The results proves that the heuristic approach has competitive features of linearity as the problem grows, but still with satisfying optimums in the evaluated instances.