A blackboard based system for context sensitive mission planning in an autonomous vehicle

This paper describes an experimental route planner for an autonomous vehicle. Given a high level mission specification (an unordered list of way-points to visit and surveys to run), the system works out a route connecting the mission tasks and issues intermediate level motion commands that describe that route. This planner is implemented in Flavors and is based on the Blackboard Control Architecture. which separates the domain specific issues from the control issues of a problem solving effort. This separation allows the planner to reason about what strategies to apply during the problem solving exercise. The control subsystem allows the planner to adopt one of two planning strategies according to an input parameter in the mission specification. The system forms a basis for context sensitive planning, which is an important issue for an autonomous vehicle. The planner is being translated into Common Lisp for installation on board an autonomous vehicle in late summer 1987.