Overview

The e-puck education robot
(photo courtesy www.e-puck.org)

The continuous operation of battery operated mobile robots requires that they have access to a reliable power source, and can access it when required. This research aimed to provide a means for an e-puck robot fitted with an inductive coil to self navigate to a contactless charging platform[1], allowing it to autonomously operate beyond the limitations of one battery cycle.

Broadly speaking, the project was undertaken in two parts. First, an object localisation strategy for an individual real-world robot was developed, modelled on the foraging behaviour of E. coli bacteria. An audio beacon was used as an artificial gradient source, such that the robot could iteravely improve its position in the environment relative to the charging platform.

After successfully demonstrating that the bacterial-inspired strategy was an effective and reliable approach to object localisation, it was extended to swarms of mobile robots in simulation. A collective navigation approach was developed based on the naturally occuring process of Diffusion Limited Aggregation (DLA), which results in a rooted tree structure that is useful for path formation and homing. The collective navigation strategy was successfully demonstrated with a simulated swarm of 60 agents, allowing agents to navigate to the charging platform with assistance from the swarm, even when outside the audible range of the audio beacon.

This research project was undertaken by Jonathan Mullins as part of the Bachelor of Software Engineering degree at Monash University, under the supervision of Dr. Bernd Meyer. The honours thesis can be downloaded here.


[1] - Jonathan Mullins, Bernd Meyer, and Patrick Hu. Collective robot navigation using diffusion limited aggregation. In International Conference on Parallel Problem Solving from Nature (PPSN-XII), Taormina, Sicily, September 2012.

[2] - Liang Jerry Chen, Weng Ip Simon, Aiguo Patrick Hu, and Bernd Meyer. A contactless charging platform for swarm robots. In Annual Conference of the IEEE Industrial Electronics Society (IECON), Melbourne, November 2011.