Measurement of Space: From Ants to Robots
Sahin Erol

Abstract:

We are interested in the ways that animals use to measure spaces that are much larger than their sensing range, and how these ways can be tested and adapted by the mobile robots. Particularly, the animals that we are concerned are the scout worker ants that assess potential nest cavities needed by their entire colony. The individual scout ants work literally in the dark and the spaces that they explore are beyond their immediate sensing range and have irregular and unpredictable shapes. Furthermore, these scouts must not only measure the floor area of a potential nest site but they must also assess the integrity of the inner periphery of the nest to check that it does not have too many holes that would make it difficult to defend against enemies. The hypothesis that the scouts use a `Buffon's needle algorithm' to estimate the nest size is supported by experiments. In this paper, we present a behavioral model for the nest size assessment of the scouts. This behavior is implemented on an {\em ant-bot}, a simulated scout model, to study the assessment process. We present the simulation results obtained from this model by systematically varying the behavior and analyzing how well the size of the nest is evaluated and how well the integrity of the periphery is checked. We also discuss future lines of research on how the robots can be used for illuminating further the behaviour of the ants, and how these studies can be useful in developing robust behaviors for autonomous mobile robots.