Conditional sequencing for land, space and sea

Conditional sequencing is a software technology for robust control of autonomous vehicles. It has been applied to terrestrial vehicles, and is currently being developed for use in spacecraft. Although it has not yet been applied to subsurface vehicles, we believe that the technology may prove useful under water as well as on land and in space. Conditional sequencing is a framework for implementing autonomous control algorithms that respond intelligently to unexpected contingencies. Because of the potential complexity of such algorithms, a key part of the technology is the use of high-level languages to express these algorithms. We are currently using Firby's RAPs language (1989), which was designed for mobile robots, and are developing a second language, CSL, whose design is tailored more towards spacecraft. The differences between RAPs and CSL are a reflection of the differences in the underlying structure of the problem in the terrestrial and space domains. For mobile robots, most of the contingencies arise from unexpected interactions between the robot and its environment. On spacecraft, most contingencies arise from hardware failures. We describe how RAPs and CSL meet the unique demands of their respective domains, and speculate on how the approach could be adapted for use under water.