We report a new method for computing near optimal makespan solutions to multi-robot path planning problem on graphs. Our focus here is with hard instances - those with up to 85% of all graph nodes occupied by robots. Our method yields 100-1000x speedup compared with existing methods. At the same time, our solutions have much smaller and often optimal makespans.

Exploring the conformational energy landscape of a molecule is an important but challenging problem because of the inherent complexity of this landscape. As part of this theme, various methods have been developed to compute transition paths between stable states of a molecule. Besides the methods classically used in biophysics/biochemistry, a recent approach originating from the robotics community has proven to be an efficient tool for conformational exploration. This approach, called the Transition-based RRT (T-RRT) is based on the combination of an effective path planning algorithm (RRT) with a Monte-Carlo-like transition test. In this paper, we propose an extension to TRRT based on a multi-tree approach, which we call Multi-T-RRT. It builds several trees rooted at different interesting points of the energy landscape and allows to quickly gain knowledge about possible conformational transition paths. We demonstrate this on the alanine dipeptide.

We provide a systematic analysis of levels of integration between discrete high-level reasoning and continuous low-level reasoning to address hybrid planning problems in robotics. We identify four distinct strategies for such an integration: (i) low-level checks are done for all possible cases in advance and then this information is used during plan generation, (ii) low-level checks are done exactly when they are needed during the search for a plan, (iii) first all plans are computed and then infeasible ones are filtered, and (iv) by means of replanning, after finding a plan, low-level checks identify whether it is infeasible or not; if it is infeasible, a new plan is computed considering the results of previous low-level checks. We perform experiments on hybrid planning problems in robotic manipulation and legged locomotion domains considering these four methods of integration, as well as some of their combinations. We analyze the usefulness of levels of integration in these domains, both from the point of view of computational efficiency (in time and space) and from the point of view of plan quality relative to its feasibility. We discuss advantages and disadvantages of each strategy in the light of experimental results and provide some guidelines on choosing proper strategies for a given domain.

Referring expressions such as “a long meeting” and “a restaurant near my brother’s” depend on information from the context to be accurately resolved. Interpreting these expressions requires pragmatic inferences that go beyond what the speaker said to what she meant; and to do this one must consider the speaker’s decisions with respect to her initial belief state and the alternative linguistic options she may have had. Modeling reference generation as a planning problem, where actions corre- spond to words that change a belief state, suggests that interpretation can also be viewed as recognizing belief- state plans that contain implicit actions. In this paper, we describe how planners can be adapted and used to interpret uncertain referring expressions.

The Traveling Tournament Problem (TTP) is a well-known NP-complete problem in sports scheduling that was inspired by the application of optimizing schedules for Major League Baseball to reduce total team travel. The techniques and heuristics from the n-team TTP can be extended to optimize the scheduling of other sports leagues, such as the Nippon Professional Baseball (NPB) league in Japan. In this paper, we describe the additional scheduling constraints required by the NPB league, such as the requirement that each team play the same number of weekend home games, weekday home games, weekend road games, and weekday road games. We fully solve this TTP-variant for the case n=6, and conclude the paper by presenting the official 2013 NPB Central League Schedule, where we helped this Japanese baseball league reduce total team travel by over six thousand kilometres.

We present GOSMR ("goal oriented scenario modeling robots"), a cognitive architecture designed to show coordinated, goal-directed behavior over the Internet, focusing on the web browser as a case study. The architecture combines a variety of artificial intelligence techniques, including planning, temporal difference learning, elementary reasoning over uncertainty, and natural language parsing, but is designed to be computationally lightweight. Its intended use is to be deployed on virtual machines in large-scale network experiments in which simulated users' adaptation in the face of resource denial should be intelligent but varied. The planning system performs temporal difference learning of action times, discounts goals according to hyperbolic discounting of time-to-completion and chance of success, takes into account the assertions of other agents, and separates abstract action from site-specific affordances. Our experiment, in which agents learn to prefer a social networking style site for sending and receiving messages, shows that utility-proportional goal selection is a reasonable alternative to Boltzmann goal selection for producing a rational mix of behavior.

We describe a model for competitive online scheduling algorithms. Two servers, each with a single observable queue, compete for customers. Upon arrival, each customer strategically chooses the queue with minimal expected wait time. Each scheduler wishes to maximize its number of customers, and can strategically select which scheduling algorithm, such as First-Come-First-Served (FCFS), to use for its queue. This induces a game played by the servers and the customers. We consider a non-Bayesian setting, where servers and customers play to maximize worst-case payoffs. We show that there is a unique subgame perfect safety-level equilibrium and we describe the associated scheduling algorithm (which is not FCFS). The uniqueness result holds for both randomized and deterministic algorithms, with a different equilibrium algorithm in each case. When the goal of the servers is to minimize competitive ratio, we prove that it is an equilibrium for each server to apply FCFS: each server obtains the optimal competitive ratio of 2.

In this paper, we study the structure and computational complexity of optimal multi-robot path planning problems on graphs. Our results encompass three formulations of the discrete multi-robot path planning problem, including a variant that allows synchronous rotations of robots along fully occupied, disjoint cycles on the graph. Allowing rotation of robots provides a more natural model for multi-robot path planning because robots can communicate. Our optimality objectives are to minimize the total arrival time, the makespan (last arrival time), and the total distance. On the structure side, we show that, in general, these objectives demonstrate a pairwise Pareto optimal structure and cannot be simultaneously optimized. On the computational complexity side, we extend previous work and show that, regardless of the underlying multi-robot path planning problem, these objectives are all intractable to compute. In particular, our NP-hardness proof for the time optimal versions, based on a minimal and direct reduction from the 3-satisfiability problem, shows that these problems remain NP-hard even when there are only two groups of robots (i.e. robots within each group are interchangeable).

The increased demand for distributed computations on “big data” has led to solutions such as SCOPE, DryadLINQ, Pig, and Hive, which allow the user to specify queries in an SQL-like language, enriched with sets of user-defined operators. The lack of exact semantics for user-defined operators interferes with the query optimization process, thus putting the burden of suggesting, at least partial, query plans on the user. In an attempt to ease this burden, we propose a formal model that allows for data-parallel program synthesis (DPPS) in a semantically well-defined manner. We show that this model generalizes existing frameworks for data-parallel computation, while providing the flexibility of query plan generation that is currently absent from these frameworks. In particular, we show how existing, off-the-shelf, AI planning tools can be used for solving DPPS tasks.

Commitments help model interactions in multiagent systems in a computationally realizable yet high-level manner without compromising the autonomy and heterogeneity of the member agents. Recent work shows how to combine commitments with goals and apply planning methods to enable agents to determine their actions. However, previous approaches to modeling commitments are confined to propositional representations, which limits their applicability in practical cases. We propose a first-order representation and reasoning technique that accommodates templatic commitments and goals that may be applied repeatedly with differing bindings for domain objects. Doing so not only leads to a more perspicuous modeling, but also supports many practical patterns.