Benchmarking bipedal walking: Towards international consensus

In the R&D community there is a growing awareness of the importance of benchmarking. Benchmarks not only allow comparing performance of different technologies, devices, and algorithms, but also define and thereby boost R&D targets and directions, and are an important support to standardization and regulation procedures towards market introduction. Benchmarking has been recently considered by the Horizon 2020 Programme the primary instrument to evaluate the Technology Readiness Levels (TRL) of the technology. The Multi-Annual Roadmap for Robotics in Europe [1] highlights the need for new benchmarks to reliably measure the performance and safety of robots. Benchmarking is gaining relevance also in the field of humanoids, as demonstrated by the success of recent competition-based initiatives (e.g Darpa Robotic Challenge, Robocup). In all these scenarios, the issue of a truly comparison with the human performance is taking a central role, since humans still represents the golden standard for interaction ability, efficiency and stability.