Parallel robot control using speculative computation

Due to the coupling in the dynamics equations, coarse-grain parallelism of robot control algorithms is particularly difficult. We have developed a new algorithm based on the Newton-Euler dynamics formulation that overcomes the serial nature of these equations, allowing a high level of parallelism. Our controller uses data from a previous control step in current calculations to allow many more tasks to be executed in parallel, thus providing higher control update rates. The use of `stale' data is an effective solution to the speedup problem, but presents some special difficulties. One stability issue when using `stale' data that is encountered in previous approaches is discussed here, along with a partial solution to the problem