Part II: Industry-driven Use-cases for Human-swarm Interaction

Project blog authored by: Mohammad Divband Soorati, Project Lead Contact | Trustworthy human-swarm partnerships in extreme environments

One of the key challenges in swarm robotics is the creation of industrial applications. Robot swarms come with a unique and valuable advantage that is fault tolerance. However, researchers in the field of swarm robotics cannot create a use-case single-handedly. In a workshop with industry experts, we co-created use-cases for human-swarm teaming that can also be used beyond our project.

 

Title: Industry Led Use-Cases for Human-Swarm Operations.

Authors: Clark, J.R., Naiseh, M., Fischer, J., Galvez Trigo, M.J., Divband Soorati, M., Bodenmann, A., Brito, M., Parnell, K. & Ramchurn, S.D. (2022). Proceedings of the AAAI 2022 Spring Symposium Series. Accepted 13 December 2021.

 

Abstract: Recent advances in technology are leading to robots of reduced size and cost. A natural outgrowth of these advances are systems comprised of large numbers of robots that collaborate autonomously in diverse applications. Research on effective autonomous control of such systems, commonly called swarms, has received attention from many domains, such as bioinspired robotics and control theory (Brambilla et al. 2013; Walker et al. 2016). Autonomous robotic swarms hold the potential to revolutionize many working domains by deploying a network of aerial, terrestrial or underwater vehicles to conduct tasks such as surveillance and payload delivery (Schranz et al., 2020). These kinds of distributed systems present novel challenges for effective Human-Swarm interaction that are only beginning to be addressed (Saffre et al., 2021). For example, when operators are tasked with operating a swarm of 50-1000 robotic agents, issues such as how the swarm can remain strategically relevant, do not violate operational boundaries and remain in a working condition arise. Underpinning these issues is the fundamental basis of how operators calibrate and maintain trust with the system (Nam et al., 2018).

Link coming soon.