Design and Density Control of a Swarm of Bimodal Particles

Abstract

In this thesis, we present the design and control of a swarm of bimodal particles that switch their geometric shape between two modes. The particles are designed and 3D printed using layers of thermoplastic polyurethane(TPU) and polylactic acid(PLA) materials and are shaped like dodecahedrons. The switching control of the geometric shape is due to the particles reaction with an external stimuli of temperature, inducing switching between a open mode classified as mode 1 and an closed mode classified as mode 2. To enact the changes in geometric shape various temperature based hot and cold programming methods were conducted using water and artificial and natural heat sources. To quantify the aggregation of the swarm and control the switching motion we utilize the metric of the Motility-Induced Phase Separation (MIPS) index. To control the particles using the MIPS index we identify the motion model of the swarm as a Brownian particle or a noisy unicycle, having different parameters for both modes. In this research we experimentally validate the noise parameters that directly affect the motion of the particles thus affecting the MIPS index, and allowing us to control the swarm more directly. Simulations in MATLAB were conducted to characterize this switching behavior using the identified noise parameters and using the identified noise parameters the aggregation of the swarm in both modes were identified. Our simulations demonstrate that with identified noise parameters that affect the particles motion, desired swarm aggregation can be achieved using simple robots that are capable of changing their geometric shape. This research highlights how the simplicity of hardware design of a single agent can achieve aggregations for swarms which enable various environmental sensing tasks to be achieved.