Page 123 - Innovations in Intelligent Machines
P. 123

114    A. Pongpunwattana and R. Rysdyk


















                           Fig. 1. The Insitu Group Seascan UAV (Top) and the Northwind Marine Seafox
                           USV (Bottom)



                                 Strategic (low bandwidth)
                                      Path Planning
                                      Task Allocation
                                      Search Patterns
                                      Human Mission Command
                                      Tactical (medium bandwidth)
                                           Target Observation
                                           Path Following
                                           Communication & Cooperation
                                           Human Monitor Interaction
                                           Dynamics and Control (high bandwidth)
                                                State Stabilization
                                                Signal Tracking
                                                Inner Loop or “autopilot”
                                                Configuration changes


                                             Fig. 2. Hierarchy of vehicle autonomy



                           Figure 2 shows three layers in the design of autonomy for air vehicles. The
                           stability and control, as well as the guidance aspects, of air vehicles are well
                           established, but strategic decision making is not. A major challenge in improv-
                           ing autonomy of unmanned vehicles is strategic planning. The focus of this
                           work is on path planning. This problem is not just planning a path between
                           two target locations, but an optimal path to visit all the targets. This is essen-
                           tially a mission planning problem for a single vehicle. The problem is composed
                           of path planning and task sequencing between various target locations.
                              There are several approaches to path planning. A popular one is graph-
                           based search. In this approach, as described by Murphy [16], the environment
   118   119   120   121   122   123   124   125   126   127   128