CONCLUSIONS Benchmarking is crucial to allow replication of results across laboratories, thus demonstrating the value of new knowledge. For empirical research there is a need to survey existing methods. In some areas of research there are established standards, subcommunities such as RoboCup have established competitions that are designed around particular benchmarks. As part of the research coordination a survey of benchmarks has been generated. So far it is too early to design new successful benchmarks, but this is an on-going effort involving more and more researchers with the aim of ensuring the wide acceptance of the community at large. The generation of the survey will no doubt increase the attention of comparative research and it is hoped that on-going work will result in the design of comprehensive benchmarks both in terms of methodology and standard data sets that will become a reference in robotics research.

Getting Involved

Attend Meetings

Communicate