Search

Robust inter-session modeling of gestures is still an open learning challenge. A sleeve equipped with capacitive strap sensors was used to capture two gesture data sets from a convenience sample of eight subjects. Two pipelines were explored. In FILT a novel two-stage algorithm was introduced which uses an unsupervised learning algorithm to find samples representing gesture transitions and discards them prior to training and validating conventional models. In TSC a confusion matrix was used to automatically consolidate commonly confused class labels, resulting in a set of gestures tailored to an individual subject’s abilities. The inter-session testing accuracy using the Time Series Consolidation (TSC) method increased from a baseline inter-session average of 42.47 ± 3.83% to 93.02% ± 4.97% while retaining an average of 5.29 ± 0.46 out of the 11 possible gesture categories. These pipelines used classic machine learning algorithms which require relatively small amounts of data and computational power compared to deep learning solutions. These methods may also offer more flexibility in interface design for users suffering from handicaps limiting their manual dexterity or ability to reliably make gestures, and be possible to implement on edge devices with low computational power.

Designers perform early-stage formative usability tests with low-fidelity prototypes to improve the design of new products. This low-tech prototype style reduces the manufacturing resources but limits the functions that can be assessed. Recent advances in technology enable designers to create low-fidelity 3D models for users to engage in a virtual environment. Three-dimensional models communicate design concepts and are not often used in formative usability testing. The proposed method discusses how to create a virtual replica of a product by assessing key human interaction steps and addresses the limitations of translating those steps into a virtual environment. In addition, the paper will provide a framework to evaluate the usability of a product in a virtual setting, with a specific emphasis on low-resource online testing in the user population. A study was performed to pilot the subject’s experience with the proposed approach and determine how the virtual online simulation impacted the performance. The study outcomes demonstrated that subjects were able to successfully interact with the virtual replica and found the simulation realistic. This method can be followed to perform formative usability tests earlier and incorporate subject feedback into future iterations of their design, which can improve safety and product efficacy.

Human variability related to physical, cognitive, socio-demographic, and other factors can contribute to large differences in human performance. Quantifying population heterogeneity can be useful for designers wishing to evaluate design parameters such that a system design is robust to this variability. Comprehensively integrating human variability in the design process poses many challenges, such as limited access to a statistically representative population and limited data collection resources. This paper discusses two virtual population modeling approaches intended to be performed prior to in-person design validation studies to minimize these challenges by: (1) targeting recruitment of representative population strata and (2) reducing the candidate design parameters being validated in the target population. The first approach suggests the use of digital human models, virtual representations of humans that can simulate system interaction to eliminate candidate design parameters. The second approach suggests the use of existing human databases to identify relevant human characteristics for representative recruitment strata in subsequent studies. Two case studies are presented to demonstrate each approach, and the benefits and limitations of each are discussed. This paper demonstrates the benefit of modeling prior to conducting in-person human performance studies to minimize resource burden, which has significant implications on early design stages.

Human interactions are paramount to the user experience, satisfaction, and risk of user errors. For products, anthropometry has traditionally been used in product sizing. However, structured methods that accurately map static and dynamic capabilities (e.g., functional mapping) of musculoskeletal regions for the conceptualization and redesign of product applications and use cases are limited. The present work aims to introduce and validate the effectiveness of the Interaction Variability method, which maps product components and musculoskeletal regions to determine explicit design parameters through limiting designer variation in the classification of human interaction factors. This study enrolled 16 engineering students to evaluate two series of interactions for (1) water bottle and (2) sunglasses applications enabling method validity and designer consistency assessments. For each interaction series, subjects identified and characterized product applications, components, and human interaction factors. Primary interactions, product mapping, and application identification achieved consensus between ranges of 31.25% and 100.00%, with significance (p < 0.1) observed at consensus rates of ≥75.00%. Significant levels of consistency were observed amongst designers, for at least one measure in all phases except anthropometric mapping for the sunglasses application indicating method effectiveness. Interaction variability was introduced and validated in this work as a standardized approach to identify, define, and map human and product interactions, which may reduce unintended use cases and user errors, respectively, in consumer populations.

Virtual environments, augmented reality, serious games, simulations, and other digitally mediated experiences are revolutionizing the way that we work, learn, and interact with people and systems. The employment of these technologies is driving innovation in research, training, education, evaluation, and various business endeavors. This session will provide a brief overview of some of the diverse uses of virtual environments (VEs) in an alternate demonstration format that leaves just over half of the session time for hands on, interactive demonstrations. Unlike most demonstration sessions, where possible, we encourage session attendees to personally interact with the demonstrations. The session will begin with demonstrators providing a brief description of their VE and how they’ve used it to address a unique need. At the conclusion of the description portion of the session, each VE will be set up at a demonstration station in the room, and session attendees are encouraged to engage with both the demonstrations for a hands on interaction with the VE and with the demonstrators to get more information about the development and use of the VE, and stimulate discussion on how the concepts, methods, and tools can be used to solve additional problems or business needs.