Decision-Theoretic Robotic Surveillance

Abstract

Acknowledgmentssviii i 66 Path-based Decisions 6.11 New clustering of our office building 6.22 The fixed cluster route strategy 6.2.11Overview 6.2.22 Candidate trajectory generation 6.2.33The decision level 6.2.44 Trajectory selection 6.33 Simulation results 6.3.11Cost-based comparison 6.3.22Sensitivity to clustering 6.44 Summary 77 Conclusions 7.11 Overview Ill 7.22 Claims 7.33 Further work 7.44 Final word AA Example POMDP application.115 A.ll Setting example 3.2.4 as a POMDP 115 A.22 One step look-ahead POMDP application A.33 Remarks related to the example Bibliographyy Indexx Samenvattingg Abstractt