Task partitioning in insects and robots. (a) Task partitioned retrieval of leaf fragments, as found in most Atta leafcutter ants that harvest leaves from trees. Dropper ants cut leaves which then accumulate in a cache, after which the leaves are retrieved by collectors and brought back to the nest, where they serve as a substrate for a fungus which is farmed as food. Ants also occasionally use a generalist strategy whereby both tasks are performed by the same individuals. (b) Analogous robotics setup, whereby items have to be transported across a slope using the coordinated action of droppers, collectors and possibly generalists. (c) Grass cutting leafcutter ants cutting leaf fragments in a flat environment without task partitioning, using a generalist foraging strategy (d) Analogous robotics setup, with robots being required to collect items in a flat arena. doi:10.1371/journal.pcbi.1004273.g001
Darwinian selection can be used to evolve robot controllers. Taking inspiration from the way in which ants, bees organize their work and divide up tasks, Eliseo Ferrante and colleagues evolved complex robot behaviors using artificial evolution and detailed robotics simulation.
The field of ‘swarm robotic’ aims to use teams of small robots to explore complex environments, such as the moon or foreign planets. However, designing controllers that allow the robots to effectively organize themselves is no easy task. The type of division of labor they set-up is known as “task partitioning”, and requires different tasks to be carried out in sequence by different sets of individuals. In particular, the experimental scenario was inspired by a spectacular form of task partitioning found in some leafcutter ants, whereby some ants (“droppers”) cut and drop leaf fragments into a temporary leaf storage cache and others (“collectors”) specialize in collecting and retrieving the fragments back to the nest.
Foot-bot robots and ARGoS simulation platform.
>> In the robotics setup, they used a team of robots simulated in-silico using an embodied swarm robotics simulator and required the robots to collect items and bring them back to the nest in either a flat or sloped environment. The results of these experiments show for the first time that complex, self-organized task specialization and task allocation could be evolved in teams of robots.
The novel method developed by the team of scientists from the University of Leuven, the Free University of Brussels and the Middle East Technical University is based on grammatical evolution and Allows the evolution of behaviours that go beyond the complexity achieved before this study.
http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1004273
https://www.youtube.com/watch?v=8mlHXcCNzjg
Recent Comments