Robots must perform tasks efficiently and reliably while acting underuncertainty. One way to achieve efficiency is to give the robot common-sense knowledge about the structure of the world. Reliable robot behaviour can be achieved by modelling the uncertaintyin the world probabilistically. We present a robot system that combines these two approaches and demonstrate the improvements in efficiency and reliability that result. Our first contribution is a probabilistic relational model integrating common-sense knowledge about the world in general, with observations of a particular environment. Our second contribution is a continual planning system which is able to plan in the large problems posed by that model, by automatically switching between decision-theoretic and classical procedures. We evaluate our system on object search tasks in two different real-world indoor environments. By reasoning about the trade-offs between possible courses of action with different informational effects, and exploiting the cues and general structures of those environments, our robot is able to consistently demonstrate efficient and reliable goal-directed behaviour.

Classical planning has been notably successful in synthesizing finite plans to achieve states where propositional goals hold. In the last few years, classical planning has also been extended to incorporate temporally extended goals, expressed in temporal logics such as LTL, to impose restrictions on the state sequences generated by finite plans. In this work, we take the next step and consider the computation of infinite plans for achieving arbitrary LTL goals. We show that infinite plans can also be obtained efficiently by calling a classical planner once over a classical planning encoding that represents and extends the composition of the planning domain and the Buchi automaton representing the goal. This compilation scheme has been implemented and a number of experiments are reported.

In today's services driven economic environment, it is imperative for organizations to provide better quality service experience to differentiate and grow their business. Customer satisfaction (C-SAT) is the key driver for retention and growth in Retail Banking. Wait time, the time spent by a customer at the branch before getting serviced, contributes significantly to C-SAT. Due to high footfall, it is improbable to improve the wait time of every customer walking in the branch. Therefore, banks in developing countries are strategically looking to segment its customers and services and offer differentiated QoS based service delivery. In this work, we present a system for customer segmentation, and scheduling based on historic value of the customer and characteristics of current service request. We describe the system and give mathematical formulation of the scheduling problem and the associated heuristics. We present results and experience of deployment of this solution in multiple branches of a leading bank in India.

The field of deterministic AI planning can roughly be divided into two approaches - classical state-based planning and hierarchical task network (HTN) planning. The plan existence problem of the former is known to be decidable while it has been proved undecidable for the latter. When extending HTN planning by allowing the unrestricted insertion of tasks and ordering constraints, one obtains a form of planning which is often referred to as "hybrid planning." We present a simplified formalization of HTN planning with and without task insertion. We show that the plan existence problem is undecidable for the HTN setting without task insertion and that it becomes decidable when allowing task insertion. In the course of the proof, we obtain an upper complexity bound of EXPSPACE for the plan existence problem for propositional HTN planning with task insertion.

Monte Carlo tree search (MCTS) methods have had recent success in games, planning, and optimization. MCTS uses results from rollouts to guide search; a rollout is a path that descends the tree with a randomized decision at each ply until reaching a leaf. MCTS results can be strongly influenced by the choice of appropriate policy to bias the rollouts. Most previous work on MCTS uses static uniform random or domain-specific policies. We describe a new MCTS method that dynamically adapts the rollout policy during search, in deterministic optimization problems. Our starting point is Cazenave's original Nested Monte Carlo Search (NMCS), but rather than navigating the tree directly we instead use gradient ascent on the rollout policy at each level of the nested search. We benchmark this new Nested Rollout Policy Adaptation (NRPA) algorithm and examine its behavior. Our test problems are instances of Crossword Puzzle Construction and Morpion Solitaire. Over moderate time scales NRPA can substantially improve search efficiency compared to NMCS, and over longer time scales NRPA improves upon all previous published solutions for the test problems. Results include a new Morpion Solitaire solution that improves upon the previous human-generated record that had stood for over 30 years.

We propose a novel approach for solving unary SAS+ planning problems. This approach extends an SAS+ instance with new state variables representing intentions about how each original state variable will be used or changed next, and splits the original actions into several stages of intention followed by eventual execution. The result is a new SAS+ instance with the same basic solutions as the original. While the transformed problem is larger, it has additional structure that can be exploited to reduce the branching factor, leading to reachable state spaces that are many orders of magnitude smaller (and hence much faster planning) in several test domains with acyclic causal graphs.

We address a difficult, yet under-investigated class of planning problems: fully-observable nondeterministic (FOND) planning problems with strong cyclic solutions. The difficulty of these strong cyclic FOND planning problems stems from the large size of the state space. Hence, to achieve efficient planning, a planner has to cope with the explosion in the size of the state space by planning along the directions that allow the goal to be reached quickly. A major challenge is: how would one know which states and search directions are relevant before the search for a solution has even begun? We first describe an NDP-motivated strong cyclic algorithm that, without addressing the above challenge, can already outperform state-of-the-art FOND planners, and then extend this NDP-motivated planner with a novel heuristic that addresses the challenge.

A plan with rich control structures like branches and loops can usually serve as a general solution that solves multiple planning instances in a domain. However, the correctness of such generalized plans is non-trivial to define and verify, especially when it comes to whether or not a plan works for all of the infinitely many instances of the problem. In this paper, we give a precise definition of a generalized plan representation called an FSA plan, with its semantics defined in the situation calculus. Based on this, we identify a class of infinite planning problems, which we call one-dimensional (1d), and prove a correctness result that 1d problems can be verified by finite means. We show that this theoretical result leads to an algorithm that does this verification practically, and a planner based on this verification algorithm efficiently generates provably correct plans for 1d problems.

The development of service robots has gained more and more attention over the last years. Advanced robots have to cope with many different situations and contingencies while executing concurrent and interruptable complex tasks. To manage the sheer variety of different execution variants the robot has to decide at run-time for the most appropriate behavior to execute. That requires task coordination mechanisms that provide the flexibility to adapt at run-time and allow to balance between alternatives.

One drawback with using plan recognition in adversarial games is that often players must commit to a plan before it is possible to infer the opponent's intentions. In such cases, it is valuable to couple plan recognition with plan repair, particularly in multi-agent domains where complete replanning is not computationally feasible. This paper presents a method for learning plan repair policies in real-time using Upper Confidence Bounds applied to Trees (UCT). We demonstrate how these policies can be coupled with plan recognition in an American football game (Rush 2008) to create an autonomous offensive team capable of responding to unexpected changes in defensive strategy. Our real-time version of UCT learns play modifications that result in a significantly higher average yardage and fewer interceptions than either the baseline game or domain-specific heuristics. Although it is possible to use the actual game simulator to measure reward offline, to execute UCT in real-time demands a different approach; here we describe two modules for reusing data from offline UCT searches to learn accurate state and reward estimators.