Hierarchical Goal Networks: Formalisms and Algorithms for Planning and Acting

In real-world applications of AI and automation such as in robotics, computer game playing and web-services, agents need to make decisions in unstructured environments that are open-world, dynamic and partially observable. In the AI and Robotics research communities in particular, there is much interest in equipping robots to operate with minimal human intervention in diverse scenarios such as in manufacturing plants, homes, hospitals, etc. Enabling agents to operate in these environments requires advanced planning and acting capabilities, some of which are not well supported by the current state of the art automated planning formalisms and algorithms. To address this problem, I propose a new planning formalism that addresses some of the inadequacies in current planning frameworks, and a suite of planning and acting algorithms that operate under this planning framework. Two parts of this work have already been accomplished. They are: • Hierarchical Goal Network (HGN) Planning. This planning formalism that combines aspects (and therefore harnesses advantages) of Classical Planning and Hierarchical Task Network (HTN) Planning, two of the most prominent planning formalisms currently in use. In particular, HGN planning algorithms, while retaining the efficiency and scalability advantages of HTNs, also allows incorporation of heuristics and other reasoning techniques from Classical Planning. • Offline Planning Algorithms. Goal Decomposition Planner (GDP) and the Goal Decomposition with Landmarks (GoDeL) planner, two HGN planning algorithms that combines hierarchical decomposition with classical planning heuristics to outperform state-of-the-art HTN planners like SHOP and SHOP2. While the work done so far relates to offline planning formalisms and algorithms, the remainder of this thesis will focus on augmenting these techniques with execution-time reasoning algorithms that enable robust execution of the generated plans. In particular, the proposed work will address two interconnected execution-time reasoning problems: • dynamically repairing plans in the face of execution failures and exogenous events; • integrating our offline HGN planners and execution-time plan repair algorithms with low-level motion and object manipulation algorithms from Robotics, thus realizing an end-to-end planning system that can be deployed onboard robots. Given the need for autonomous agents to operate in open, dynamic and unstructured environments and the obvious need for high-level deliberation capabilities to enable intelligent behavior, we believe the planning-and-acting systems that are developed as part of this thesis could provide unique insights into ways to realize these systems in the real world.