Pure Vision-Based Motion Tracking for Data-Driven Design – A Simple, Flexible, and Cost-Effective Approach for Capturing Static and Dynamic Interactions

Abstract

This paper presents an exploratory case study where video-based pose estimation is used to analyse human motion to support data-driven design. It provides two example use cases related to design. Results are compared to ground truth measurements showing high correlation for the estimated pose, with an RMSE of 65.5 mm. The paper exemplifies how design projects can benefit from a simple, flexible, and cost-effective approach to capture human-object interactions. This also entails the possibility of implementing interaction and body capturing in the earliest stages of design, at minimal effort.

Human-centred engineering design: a cross-disciplinary product innovation practice

“…Integrating sensor technology with human-centred design practices enables designers to create experiments that measure various humanproduct interactions. Open-source sensor systems capable of reading dynamic motions are rapidly developing (Johnston et al, 2022). These developments provide new opportunities, enabling integrating various design practices to re-think the design of existing products.…”

“…For example, motion capture is a known procedure to capture movements, aiding design of equipment (Severin et al, 2021). The development of motion caption systems from reflector-based to video-based enables designers to perform in-situ experiments outside laboratory environments (Johnston et al, 2022). Analysis of geometry: Gaining insight into human-product interactions in a dynamic setting enable designers to understand all of the different interactions.…”

“…With a set product design, we explored possible redesigns of oars and rowlocks in an experiment because it has the potential to further increase athlete performance, and validate the final design in its intended use context. The in-situ-i.e., on water-experiment compared three rowlock positions, and collected data from multiple sensors, including motion capture (Johnston et al, 2022). Previous baseline experiments had been performed both in lab and on-water, allowing us to compare lab and in-situ motion envelope and performance (produced power).…”

A low-cost non-intrusive spatial hand tracking pipeline for product-process interaction

“…In products, mass-customisation is pushing for individualised and tailored experiences requiring designers to have a deeper and better understanding of how users interact with their product. Human factors research has shown that supplementing traditionally qualitative methodologies with high-rate quantitative data inputs both reduces design cycle time and allows for improved product design (Johnston et al, 2022). It is not only important to capture track hands spatially and to a high degree of accuracy but to also classify what the hands and/or activity the individual is doing when interacting with the product (Joong Hee Lee & Yun, 2023;Shi et al, 2021).…”

People, space use and objects: an UWB-based quantifying approach for post-occupancy evaluation of new architectural spaces

“…Alternatively, wearable sensors can be used, either readily available smartphones (Kostakos et al ., 2013; Kotanen et al ., 2003; Tiku et al ., 2023; Tiku and Pasricha, 2023) or custom wearable devices (Sjöman and Steinert, 2016; Smith et al ., 2013). Finally, computer vision can be used for data analysis as well as manual observation (Abbas et al ., 2012), but this approach does present privacy concerns or lack scalability (Johnston et al ., 2022). Each approach provides data that can be used for either detection, counting or tracking occupants with suitable algorithms (Saha et al ., 2019).…”