Berke A., Nawyn J., Lengeling T.S., Larson K. (2020) Urban Mobility Swarms: A Scalable Implementation. In: Arai K., Kapoor S., Bhatia R. (eds) Intelligent Computing. SAI 2020. Advances in Intelligent Systems and Computing, vol 1229. Springer, Cham
Work for a Member company and need a Member Portal account? Register here with your company email address.
Berke A., Nawyn J., Lengeling T.S., Larson K. (2020) Urban Mobility Swarms: A Scalable Implementation. In: Arai K., Kapoor S., Bhatia R. (eds) Intelligent Computing. SAI 2020. Advances in Intelligent Systems and Computing, vol 1229. Springer, Cham
We present a system to coordinate “urban mobility swarms” in order to promote the use and safety of lightweight, sustainable transit, while enhancing the vibrancy and community fabric of cities. This work draws from behavior exhibited by swarms of nocturnal insects, such as crickets and fireflies, whereby synchrony unifies individuals in a decentralized network. Coordination naturally emerges in these cases and provides a compelling demonstration of “strength in numbers”. Our work is applied to coordinating lightweight vehicles, such as bicycles, which are automatically inducted into ad-hoc “swarms”, united by the synchronous pulsation of light. We model individual riders as nodes in a decentralized network and synchronize their behavior via a peer-to-peer message protocol and algorithm, which preserves individual privacy. Nodes broadcast over radio with a transmission range tuned to localize swarm membership. Nodes then join or disconnect from others based on proximity, accommodating the dynamically changing topology of urban mobility networks. This paper provides a technical description of our system, including the protocol and algorithm to coordinate the swarming behavior that emerges from it. We also demonstrate its implementation in code, circuity, and hardware, with a system prototype tested on a city bike-share. In doing so, we evince the scalability of our system. Our prototype uses low-cost components, and bike-share programs, which manage bicycle fleets distributed across cities, could deploy the system at city-scale. Our flexible, decentralized design allows additional bikes to then connect with the network, enhancing its scale and impact.
This paper is available via arXiv: https://arxiv.org/pdf/2007.06653.pdf