Knowledge Intensive Mixed Initiative Mission Planning Technologies for Cooperative Autonomous Systems

Purpose and goal

The objective of the project has been to develop mission planning methods that can handle for example temporal uncertainty and mixed initiative (operator/vehicle) with knowledge support from an ontology. The original objectives of the planning work package (planning for actions with temporal uncertainty, resource usage, planning domain) have been extended with new mission types (surveillance), external schedules, and fuel optimization. The mixed initiative and ontology work packages have been reduced in scope to compensate for the extensions in planning.

Expected results and effects

The new algorithms support planning / interaction for a wide class of missions that previously had to be managed by hand, which results in large time and quality gains. Significant emphasis has been placed on developing methods that are not narrowly limited to the current problem description but directly support future extensions. This provides an excellent basis for future research. The project has contributed with ideas that have been used in the preparation of the research program WASP, Wallenberg Autonomous Systems and Software Program.

Planned approach and implementation

The project has been carried out as a series of iterations: New theory leads to prototype development, which provides feedback to the next level of theory. This proven method has worked very well and given Saab significant opportunities to (re)prioritize areas of development during the project. In some periods, tighter collaboration has been possible through a Saab employee who has worked at LiU one day per week. Continuing this process during the last year would have been valuable, but this was unfortunately not possible.