Internship about Human-Robot Interaction with the HMI Group of University of Twente
For three months I worked on improving handovers between robots and people with blindness. This included a self-ethnography study about my handovers in everyday life, interviews with users with blindness, and finally a user test with participants with blindness.The Problem
Generally, when robots hand over objects to people with blindness, they face issues like lack of gaze coordination, lack of understanding the need of the user, and a communication gap. This leads to problems where a person with blindness, while receiving an object from a robot might accidentally drop the object or have a harder time locating it. This inturn will lead to problems in everyday life, when robots will start prevailing in all walks of life.Solution
The solution to above said problem would be to understand the diverse needs of a person with blindness. What do they have problems with in everyday life with other people, do they have issues with direct handovers or indirect handovers, how does it differ for different objects and scenarios. understanding these factors would help in building a handover system that works for everyone, regardless of sight.Research Methodology
For the research, I first started reading the literature on human-robot interaction and how people with blindness are included in it. While I found quite a few papers that spoke about robotic navigation, and other related topics, there were very few papers that researched handovers between robots and people with blindness (( (Käppler et al. (2021); Langer, et al. (2024).
This lead to a self-ethnography study on my experience with handovers on a day-today basis. I recorded my handovers for nearly three weeks. I logged the date, the time, the type of handover, type of object, familiar/unfamiliar person, handover state, and location. Later, I analysed the results with thematic analysis.
For the next phase of the research, I interviewed 10 participants with blindness to understand how handovers worked in their everyday life with other people, Where I tried to understand the different variation in handovers, different techniques that the participants used, and the level of verbal guidance they required for the handovers.
for the final research phase, I collaborated with a student robotics engineer to build an audio-based system that helps in locating the arm of the robot using different variations of beeps. We tested the system through a robotic arm with 3 participant with blindness.
Currently, I am analysing the outcome of our user study
References
Käppler, M., Deml, B., Stein, T., Nagl, J., & Steingrebe, H. (2020). The importance of feedback for object hand-overs between human and robot. In T. Ahram, R. Taiar, K. Langlois, & A. Choplin (Eds.), Human Interaction, Emerging Technologies and Future Applications III (pp. 29–35). Springer. https://doi.org/10.1007/978-3-030-55307-4_5
Langer, D., Legler, F., Diekmann, P., Dettmann, A., Glende, S., & Bullinger, A. C. (2024). Got it? Comparative ergonomic evaluation of robotic object handover for visually impaired and sighted users. Robotics, 13(3), Article 43. https://doi.org/10.3390/robotics13030043