Evaluating the effect of robot group size on relative localisation precision

Looking on co-operative position estimation in multi-robot systems, the question to what extend the number of robots has an influence on the quality of the resulting localisation is an important and interesting issue. This paper addresses this relation regarding a pure relative localisation approach based only on mutual observations between the robots. The intuitive expectation that more robots should improve the position estimation is motivated and the design of the experiments with special respect to possibly distorting parameters is discussed and reasoned in detail. An in-depth analysis of the collected data explains the only partial conformance of the experimental results with the expected outcome