Autonomous Task Partitioning in Swarms of Robots: an Approach Based on Cost Estimation

The information provided is the sole responsibility of the authors and does not necessarily reflect the opinion of the members of IRIDIA. The authors take full responsibility for any copyright breaches that may result from publication of this paper in the IRIDIA – Technical Report Series. IRIDIA is not responsible for any use that might be made of data appearing in this publication. In this work, we study task partitioning in swarms of robots. Task partitioning is the process by which a task is decomposed into sub-tasks. We consider here the case in which a given task requires that a total amount of work is performed, and each sub-task incrementally contributes a fraction of that amount of work. We propose an approach that enables a swarm of robots to autonomously partition a task. Each robot performs a sub-task; the entity of the amount of work that the sub-task contributes to the execution of the overall task is decided autonomously by the robot. The entity of the contribution has an impact on the cost of performing the overall task: in some cases, partitioning the task in many small sub-tasks may be more advantageous than partitioning it in few larger sub-tasks, or vice versa. Each robot decides the entity of the amount of work that it contributes with its sub-task on the basis of a model that it builds. This model provides an estimate of the cost of performing the overall task as a function of the amount of work contributed by a sub-task, which is an indicator of the number of sub-tasks into which the original task is partitioned. We test our approach on a foraging scenario: the robots collect objects from a source location, and transport them to a target location. The robots use odometry to be able to move back to the source location. Odometry is appealing due to its simplicity and the fact that it does not require external infrastructures. However, it suffers from the accumulation of errors in time. Task partitioning can be used to reduce the effects of the odometry error. A robot transporting an object travels a limited distance before depositing the object on the ground. There, another robot picks it up and continues transportation. In such context , autonomous task partitioning consists in deciding the distance traveled by a robot when carrying an object, that is, the amount of work that the robot contributes …