Integrating Reinforcement Learning, Action Model Learning, and Numeric Planning for Tackling Complex Tasks

Automated Planning algorithms require a model of the domain that specifies the preconditions and effects of each action. Obtaining such a domain model is notoriously hard. Algorithms for learning domain models exist, yet it remains unclear whether learning a domain model and planning is an effective approach for numeric planning environments, i.e., where states include discrete and numeric state variables. In this work, we explore the benefits of learning a numeric domain model and compare it with alternative model-free solutions. As a case study, we use two tasks in Minecraft, a popular sandbox game that has been used as an AI challenge. First, we consider an offline learning setting, where a set of expert trajectories are available to learn from. This is the standard setting for learning domain models. We used the Numeric Safe Action Model Learning (NSAM) algorithm to learn a numeric domain model and solve new problems with the learned domain model and a numeric planner. We call this model-based solution NSAM_(+p), and compare it to several model-free Imitation Learning (IL) and Offline Reinforcement Learning (RL) algorithms. Empirical results show that some IL algorithms can learn faster to solve simple tasks, while NSAM_(+p) allows solving tasks that require long-term planning and enables generalizing to solve problems in larger environments. Then, we consider an online learning setting, where learning is done by moving an agent in the environment. For this setting, we introduce RAMP. In RAMP, observations collected during the agent's execution are used to simultaneously train an RL policy and learn a planning domain action model. This forms a positive feedback loop between the RL policy and the learned domain model. We demonstrate experimentally the benefits of using RAMP, showing that it finds more efficient plans and solves more problems than several RL baselines.

Transient Multi-Agent Path Finding for Lifelong Navigation in Dense Environments

Multi-Agent Path Finding (MAPF) deals with finding conflict-free paths for a set of agents from an initial configuration to a given target configuration. The Lifelong MAPF (LMAPF) problem is a well-studied online version of MAPF in which an agent receives a new target when it reaches its current target. The common approach for solving LMAPF is to treat it as a sequence of MAPF problems, periodically replanning from the agents' current configurations to their current targets. A significant drawback in this approach is that in MAPF the agents must reach a configuration in which all agents are at their targets simultaneously, which is needlessly restrictive for LMAPF. Techniques have been proposed to indirectly mitigate this drawback. We describe cases where these mitigation techniques fail. As an alternative, we propose to solve LMAPF problems by solving a sequence of modified MAPF problems, in which the objective is for each agent to eventually visit its target, but not necessarily for all agents to do so simultaneously. We refer to this MAPF variant as Transient MAPF (TMAPF) and propose several algorithms for solving it based on existing MAPF algorithms. A limited experimental evaluation identifies some cases where using a TMAPF algorithm instead of a MAPF algorithm with an LMAPF framework can improve the system throughput significantly.

Algorithm Selection for Optimal Multi-Agent Path Finding via Graph Embedding

Multi-agent path finding (MAPF) is the problem of finding paths for multiple agents such that they do not collide. This problem manifests in numerous real-world applications such as controlling transportation robots in automated warehouses, moving characters in video games, and coordinating self-driving cars in intersections. Finding optimal solutions to MAPF is NP-Hard, yet modern optimal solvers can scale to hundreds of agents and even thousands in some cases. Different solvers employ different approaches, and there is no single state-of-the-art approach for all problems. Furthermore, there are no clear, provable, guidelines for choosing when each optimal MAPF solver to use. Prior work employed Algorithm Selection (AS) techniques to learn such guidelines from past data. A major challenge when employing AS for choosing an optimal MAPF algorithm is how to encode the given MAPF problem. Prior work either used hand-crafted features or an image representation of the problem. We explore graph-based encodings of the MAPF problem and show how they can be used on-the-fly with a modern graph embedding algorithm called FEATHER. Then, we show how this encoding can be effectively joined with existing encodings, resulting in a novel AS method we call MAPF Algorithm selection via Graph embedding (MAG). An extensive experimental evaluation of MAG on several MAPF algorithm selection tasks reveals that it is either on-par or significantly better than existing methods.

Optimal and Bounded Suboptimal Any-Angle Multi-agent Pathfinding

Multi-agent pathfinding (MAPF) is the problem of finding a set of conflict-free paths for a set of agents. Typically, the agents' moves are limited to a pre-defined graph of possible locations and allowed transitions between them, e.g. a 4-neighborhood grid. We explore how to solve MAPF problems when each agent can move between any pair of possible locations as long as traversing the line segment connecting them does not lead to the collision with the obstacles. This is known as any-angle pathfinding. We present the first optimal any-angle multi-agent pathfinding algorithm. Our planner is based on the Continuous Conflict-based Search (CCBS) algorithm and an optimal any-angle variant of the Safe Interval Path Planning (TO-AA-SIPP). The straightforward combination of those, however, scales poorly since any-angle path finding induces search trees with a very large branching factor. To mitigate this, we adapt two techniques from classical MAPF to the any-angle setting, namely Disjoint Splitting and Multi-Constraints. Experimental results on different combinations of these techniques show they enable solving over 30% more problems than the vanilla combination of CCBS and TO-AA-SIPP. In addition, we present a bounded-suboptimal variant of our algorithm, that enables trading runtime for solution cost in a controlled manner.

Safe Learning of PDDL Domains with Conditional Effects -- Extended Version

Powerful domain-independent planners have been developed to solve various types of planning problems. These planners often require a model of the acting agent's actions, given in some planning domain description language. Manually designing such an action model is a notoriously challenging task. An alternative is to automatically learn action models from observation. Such an action model is called safe if every plan created with it is consistent with the real, unknown action model. Algorithms for learning such safe action models exist, yet they cannot handle domains with conditional or universal effects, which are common constructs in many planning problems. We prove that learning non-trivial safe action models with conditional effects may require an exponential number of samples. Then, we identify reasonable assumptions under which such learning is tractable and propose SAM Learning of Conditional Effects (Conditional-SAM), the first algorithm capable of doing so. We analyze Conditional-SAM theoretically and evaluate it experimentally. Our results show that the action models learned by Conditional-SAM can be used to solve perfectly most of the test set problems in most of the experimented domains.

Enhancing Numeric-SAM for Learning with Few Observations

A significant challenge in applying planning technology to real-world problems lies in obtaining a planning model that accurately represents the problem's dynamics. Numeric Safe Action Models Learning (N-SAM) is a recently proposed algorithm that addresses this challenge. It is an algorithm designed to learn the preconditions and effects of actions from observations in domains that may involve both discrete and continuous state variables. N-SAM has several attractive properties. It runs in polynomial time and is guaranteed to output an action model that is safe, in the sense that plans generated by it are applicable and will achieve their intended goals. To preserve this safety guarantee, N-SAM must observe a substantial number of examples for each action before it is included in the learned action model. We address this limitation of N-SAM and propose N-SAM*, an enhanced version of N-SAM that always returns an action model where every observed action is applicable at least in some state, even if it was only observed once. N-SAM* does so without compromising the safety of the returned action model. We prove that N-SAM* is optimal in terms of sample complexity compared to any other algorithm that guarantees safety. An empirical study on a set of benchmark domains shows that the action models returned by N-SAM* enable solving significantly more problems compared to the action models returned by N-SAM.

A Domain-Independent Agent Architecture for Adaptive Operation in Evolving Open Worlds

Model-based reasoning agents are ill-equipped to act in novel situations in which their model of the environment no longer sufficiently represents the world. We propose HYDRA - a framework for designing model-based agents operating in mixed discrete-continuous worlds, that can autonomously detect when the environment has evolved from its canonical setup, understand how it has evolved, and adapt the agents' models to perform effectively. HYDRA is based upon PDDL+, a rich modeling language for planning in mixed, discrete-continuous environments. It augments the planning module with visual reasoning, task selection, and action execution modules for closed-loop interaction with complex environments. HYDRA implements a novel meta-reasoning process that enables the agent to monitor its own behavior from a variety of aspects. The process employs a diverse set of computational methods to maintain expectations about the agent's own behavior in an environment. Divergences from those expectations are useful in detecting when the environment has evolved and identifying opportunities to adapt the underlying models. HYDRA builds upon ideas from diagnosis and repair and uses a heuristics-guided search over model changes such that they become competent in novel conditions. The HYDRA framework has been used to implement novelty-aware agents for three diverse domains - CartPole++ (a higher dimension variant of a classic control problem), Science Birds (an IJCAI competition problem), and PogoStick (a specific problem domain in Minecraft). We report empirical observations from these domains to demonstrate the efficacy of various components in the novelty meta-reasoning process.

Heuristic Search For Physics-Based Problems: Angry Birds in PDDL+

This paper studies how a domain-independent planner and combinatorial search can be employed to play Angry Birds, a well established AI challenge problem. To model the game, we use PDDL+, a planning language for mixed discrete/continuous domains that supports durative processes and exogenous events. The paper describes the model and identifies key design decisions that reduce the problem complexity. In addition, we propose several domain-specific enhancements including heuristics and a search technique similar to preferred operators. Together, they alleviate the complexity of combinatorial search. We evaluate our approach by comparing its performance with dedicated domain-specific solvers on a range of Angry Birds levels. The results show that our performance is on par with these domain-specific approaches in most levels, even without using our domain-specific search enhancements.

Learning to Operate in Open Worlds by Adapting Planning Models

Planning agents are ill-equipped to act in novel situations in which their domain model no longer accurately represents the world. We introduce an approach for such agents operating in open worlds that detects the presence of novelties and effectively adapts their domain models and consequent action selection. It uses observations of action execution and measures their divergence from what is expected, according to the environment model, to infer existence of a novelty. Then, it revises the model through a heuristics-guided search over model changes. We report empirical evaluations on the CartPole problem, a standard Reinforcement Learning (RL) benchmark. The results show that our approach can deal with a class of novelties very quickly and in an interpretable fashion.

An Example of the SAM+ Algorithm for Learning Action Models for Stochastic Worlds

In this technical report, we provide a complete example of running the SAM+ algorithm, an algorithm for learning stochastic planning action models, on a simplified PPDDL version of the Coffee problem. We provide a very brief description of the SAM+ algorithm and detailed description of our simplified version of the Coffee domain, and then describe the results of running it on the simplified Coffee domain.