RemoteBunnies: Multi-Agent Data Mapping Between Physical Environments

This thesis describes a method for approximating the locomotive behaviors of wild animals using near real-time robotic simulations of those behaviors. The method is realized in two subsystems. The capture subsystem uses a network of distributed RFID readers that respond to signals from active tags carried by the individuals. The chief innovation of the capture system is the design of an algorithm that uses tag-specific signal strength thresholds to determine approximate tag proximity. Software was designed to translate those "presence" signals into streams of arrival and departure events. These are then used by a representational subsystem that assigns robotic avatars to individual tag-holders. As individuals move among the readers, the robots follow their "paths" at the discrete granularity dictated by the number of readers available; the subsystem includes a rudimentary collision avoidance mechanism. No attempt is made to model behaviors between departure and arrival events. The main goal of the work presented in this document was to achieve full connectivity between the two proposed subsystems that conform RemoteBunnies. RemoteBunnies components could support applications in a variety of applications domains. As a reference, this document covers the design and implementation of RemoteBunnies as a proof-of-concept for supporting classroom activities around the foraging behavior of cottontail rabbits.