Hearing the shape of an arena with spectral swarm robotics - Supplementary Information videos

Supplementary Information Videos of the article "Hearing the shape of an arena with spectral swarm robotics".Video SI1 presents the results of the expe-2 case, involving experiments with 25 robots in arenas of surface equals to S = 70000mm2. First, we display the observed behavior of the algorithm on one experimental run with the disk and annulus arenas. This is illustrated by a time-lapse showing the active LED colors of each robots, both in the “diffusion” and “λ2 estimation” stages of the algorithm. Then, we concurrently present the results obtained through all experimental runs of Supplementary Table S6.Video SI2 compares experimental results of case expe-2 with simulation results of case sim-2. Both experimental and simulated agents are driven by the same controllers implementing the spectral swarm robotics algorithm on a Kilobot architecture. In either cases, we observe similar dynamics, for the “diffusion” (swarm exhibit a correct partitioning) and “λ2 estimation” (swarm correctly classify arenas) stages.Video SI3 showcases swarm behavior in the sim-large-7 case, with parameters of regime r3 (shown in Main Text Figs. 4 and 5). It involves 300 agents with a field of perception of 85mm and 7 arenas of surface equals to S = 500000mm2. This video presents a time lapse similar to the one in Main Text Fig. 4, both for the “diffusion” and “λ2 estimation” stages of the algorithm.Video SI4 presents the results of the pogo-expe-2 case, involving experiments with 25 Pogobot robots in arenas of surface area S = 150000 mm2. First, we display the observed behavior of the algorithm on one experimental run for both the disk and annulus arenas. This is illustrated by a time-lapse showing the active LED colors of each robot during both the diffusio and λ2 estimation stages of the algorithm. We then present the results obtained across all Pogobot experimental runs.