It is often beneficial for an autonomous agent that operates in a complex environment to make use of different types of mathematical models to keep track of unobservable parts of the world or to perform prediction, planning and other types of reasoning. There always exists a tradeoff between the model's accuracy and feasibility when it is used within a certain application, due to the limited available computational resources. In most cases, this tradeoff is to a large extent balanced by humans for model construction in general and for autonomous agents in particular.
The DARE project has investigated different solutions where agents are more responsible for balancing the tradeoff for models themselves, in the context of interleaved task planning and plan execution. The necessary components for an autonomous agent that performs its abstractions and constructs planning models dynamically during task planning and execution have been investigated. The DARE method is a template for handling the possible situations that can occur such as the rise of unsuitable abstractions and need for dynamic construction of abstraction levels. Implementations of DARE have been presented in two case studies where both a fully and partially observable stochastic domain are used, motivated by research with Unmanned Aircraft Systems. The case studies also demonstrate possible ways to perform dynamic abstraction and problem model construction in practice.
|||Per Nyblom and Patrick Doherty. Towards Automatic Model Generation by Optimization. In Proceedings of the 10th Scandinavian Conference on Artificial Intelligence (SCAI), volume 173 of Frontiers in Artificial Intelligence and Applications, Stockholm, May 2008. IOS Press, Amsterdam, The Netherlands. [ Conference | .pdf ]|
|||Per Nyblom. Dynamic Abstraction for Interleaved Task Planning and Execution. Licentiate thesis, Linköping University, Department of Computer and Information Science, May 2008. Report code: LiU-Tek-Lic-2008:21. [ E-Press | .pdf ]|
|||Per Nyblom. Dynamic Planning Problem Generation in a UAV Domain. In Proceedings of the 6th IFAC Symposium on Intelligent Autonomous Vehicles (IAV), Toulouse, France, September 2007. [ Conference | .pdf ]|
|||Per Nyblom. Dynamic Abstraction for Hierarchical Problem Solving and Execution in Stochastic Dynamic Environments. In Loris Penserini, Pavlos Peppas, and Anna Perini, editors, Proceedings of the 3rd European Starting AI Researcher Symposium (STAIRS), volume 142 of Frontiers in Artificial Intelligence and Applications, pages 263-264, Riva del Garda, Italy, August 2006. IOS Press, Amsterdam, The Netherlands. [ .pdf ]|
|||Per Nyblom. Handling Uncertainty by Interleaving Cost-Aware Classical Planning with Execution. In Peter Funk, Thorsteinn Rögnvaldsson, and Ning Xiong, editors, Proceedings of the 3rd Joint Workshop of the Swedish Artificial Intelligence Society and the Swedish Society for Learning Systems (SAIS-SSLS), pages 134-140, April 2005. [ Conference | .pdf ]|