PIBT is a rule-based Multi-Agent Path Finding (MAPF) solver, widely used as a low-level planner or action sampler in many state-of-the-art approaches. Its primary advantage lies in its exceptional speed, enabling action selection for thousands of agents within milliseconds by considering only the immediate next timestep. However, this short-horizon design leads to poor performance in scenarios where agents have orientation and must perform time-consuming rotation actions. In this work, we present an enhanced version of PIBT that addresses this limitation by incorporating multi-action operations. We detail the modifications introduced to improve PIBT's performance while preserving its hallmark efficiency. Furthermore, we demonstrate how our method, when combined with graph guidance technique and large neighborhood search optimization, achieves state-of-the-art performance in the online LMAPF-T setting.

Multi-agent pathfinding (MAPF) is a problem that generally requires finding collision-free paths for multiple agents in a shared environment. Solving MAPF optimally, even under restrictive assumptions, is NP-hard, yet efficient solutions for this problem are critical for numerous applications, such as automated warehouses and transportation systems. Recently, learning-based approaches to MAPF have gained attention, particularly those leveraging deep reinforcement learning. Typically, such learning-based MAPF solvers are augmented with additional components like single-agent planning or communication. Orthogonally, in this work we rely solely on imitation learning that leverages a large dataset of expert MAPF solutions and transformer-based neural network to create a foundation model for MAPF called MAPF-GPT. The latter is capable of generating actions without additional heuristics or communication. MAPF-GPT demonstrates zero-shot learning abilities when solving the MAPF problems that are not present in the training dataset. We show that MAPF-GPT notably outperforms the current best-performing learnable MAPF solvers on a diverse range of problem instances and is computationally efficient during inference.

Multi-agent reinforcement learning (MARL) has recently excelled in solving challenging cooperative and competitive multi-agent problems in various environments, typically involving a small number of agents and full observability. Moreover, a range of crucial robotics-related tasks, such as multi-robot pathfinding, which have traditionally been approached with classical non-learnable methods (e.g., heuristic search), are now being suggested for solution using learning-based or hybrid methods. However, in this domain, it remains difficult, if not impossible, to conduct a fair comparison between classical, learning-based, and hybrid approaches due to the lack of a unified framework that supports both learning and evaluation. To address this, we introduce POGEMA, a comprehensive set of tools that includes a fast environment for learning, a problem instance generator, a collection of predefined problem instances, a visualization toolkit, and a benchmarking tool for automated evaluation. We also introduce and define an evaluation protocol that specifies a range of domain-related metrics, computed based on primary evaluation indicators (such as success rate and path length), enabling a fair multi-fold comparison. The results of this comparison, which involves a variety of state-of-the-art MARL, search-based, and hybrid methods, are presented.

Multi-agent pathfinding (MAPF) is a common abstraction of multi-robot trajectory planning problems, where multiple homogeneous robots simultaneously move in the shared environment. While solving MAPF optimally has been proven to be NP-hard, scalable, and efficient, solvers are vital for real-world applications like logistics, search-and-rescue, etc. To this end, decentralized suboptimal MAPF solvers that leverage machine learning have come on stage. Building on the success of the recently introduced MAPF-GPT, a pure imitation learning solver, we introduce MAPF-GPT-DDG. This novel approach effectively fine-tunes the pre-trained MAPF model using centralized expert data. Leveraging a novel delta-data generation mechanism, MAPF-GPT-DDG accelerates training while significantly improving performance at test time. Our experiments demonstrate that MAPF-GPT-DDG surpasses all existing learning-based MAPF solvers, including the original MAPF-GPT, regarding solution quality across many testing scenarios. Remarkably, it can work with MAPF instances involving up to 1 million agents in a single environment, setting a new milestone for scalability in MAPF domains.

Multi-agent Pathfinding (MAPF) problem generally asks to find a set of conflict-free paths for a set of agents confined to a graph and is typically solved in a centralized fashion. Conversely, in this work, we investigate the decentralized MAPF setting, when the central controller that possesses all the information on the agents' locations and goals is absent and the agents have to sequentially decide the actions on their own without having access to the full state of the environment. We focus on the practically important lifelong variant of MAPF, which involves continuously assigning new goals to the agents upon arrival to the previous ones. To address this complex problem, we propose a method that integrates two complementary approaches: planning with heuristic search and reinforcement learning through policy optimization. Planning is utilized to construct and re-plan individual paths. We enhance our planning algorithm with a dedicated technique tailored to avoid congestion and increase the throughput of the system. We employ reinforcement learning to discover the collision avoidance policies that effectively guide the agents along the paths. The policy is implemented as a neural network and is effectively trained without any reward-shaping or external guidance. We evaluate our method on a wide range of setups comparing it to the state-of-the-art solvers. The results show that our method consistently outperforms the learnable competitors, showing higher throughput and better ability to generalize to the maps that were unseen at the training stage. Moreover our solver outperforms a rule-based one in terms of throughput and is an order of magnitude faster than a state-of-the-art search-based solver. The code is available at https://github.com/AIRI-Institute/learn-to-follow.

The Multi-Agent Pathfinding (MAPF) problem involves finding a set of conflict-free paths for a group of agents confined to a graph. In typical MAPF scenarios, the graph and the agents' starting and ending vertices are known beforehand, allowing the use of centralized planning algorithms. However, in this study, we focus on the decentralized MAPF setting, where the agents may observe the other agents only locally and are restricted in communications with each other. Specifically, we investigate the lifelong variant of MAPF, where new goals are continually assigned to the agents upon completion of previous ones. Drawing inspiration from the successful AlphaZero approach, we propose a decentralized multi-agent Monte Carlo Tree Search (MCTS) method for MAPF tasks. Our approach utilizes the agent's observations to recreate the intrinsic Markov decision process, which is then used for planning with a tailored for multi-agent tasks version of neural MCTS. The experimental results show that our approach outperforms state-of-the-art learnable MAPF solvers. The source code is available at https://github.com/AIRI-Institute/mats-lp.

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 a 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.

Multi-agent pathfinding (MAPF) is the problem of finding a set of conflict-free paths for a set of agents. 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 a 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. 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.

Multi-agent pathfinding (MAPF) is a problem that involves finding a set of non-conflicting paths for a set of agents confined to a graph. In this work, we study a MAPF setting, where the environment is only partially observable for each agent, i.e., an agent observes the obstacles and other agents only within a limited field-of-view. Moreover, we assume that the agents do not communicate and do not share knowledge on their goals, intended actions, etc. The task is to construct a policy that maps the agent's observations to actions. Our contribution is multifold. First, we propose two novel policies for solving partially observable MAPF (PO-MAPF): one based on heuristic search and another one based on reinforcement learning (RL). Next, we introduce a mixed policy that is based on switching between the two. We suggest three different switch scenarios: the heuristic, the deterministic, and the learnable one. A thorough empirical evaluation of all the proposed policies in a variety of setups shows that the mixing policy demonstrates the best performance is able to generalize well to the unseen maps and problem instances, and, additionally, outperforms the state-of-the-art counterparts (PRIMAL2 and PICO). The source-code is available at https://github.com/AIRI-Institute/when-to-switch.

Heuristic search algorithms, e.g. A*, are the commonly used tools for pathfinding on grids, i.e. graphs of regular structure that are widely employed to represent environments in robotics, video games, etc. Instance-independent heuristics for grid graphs, e.g. Manhattan distance, do not take the obstacles into account, and thus the search led by such heuristics performs poorly in obstacle-rich environments. To this end, we suggest learning the instance-dependent heuristic proxies that are supposed to notably increase the efficiency of the search. The first heuristic proxy we suggest to learn is the correction factor, i.e. the ratio between the instance-independent cost-to-go estimate and the perfect one (computed offline at the training phase). Unlike learning the absolute values of the cost-to-go heuristic function, which was known before, learning the correction factor utilizes the knowledge of the instance-independent heuristic. The second heuristic proxy is the path probability, which indicates how likely the grid cell is lying on the shortest path. This heuristic can be employed in the Focal Search framework as the secondary heuristic, allowing us to preserve the guarantees on the bounded sub-optimality of the solution. We learn both suggested heuristics in a supervised fashion with the state-of-the-art neural networks containing attention blocks (transformers). We conduct a thorough empirical evaluation on a comprehensive dataset of planning tasks, showing that the suggested techniques i) reduce the computational effort of the A* up to a factor of 4x while producing the solutions, whose costs exceed those of the optimal solutions by less than 0.3% on average; ii) outperform the competitors, which include the conventional techniques from the heuristic search, i.e. weighted A*, as well as the state-of-the-art learnable planners. The project web-page is: https://airi-institute.github.io/TransPath/.

The Multi-agent Multi-item Pickup and Delivery problem (MAMPD) stands for a problem of finding collision-free trajectories for a fleet of mobile agents transporting a set of items from their initial positions to specified locations. Each agent can carry multiple items up to a given capacity. We study the solution quality of the naive decoupled approach, which decouples the problem into task assignment (TA) and Multi-Agent Pathfinding (MAPF). By computing the gap between the lower bound of the MAMPD cost, estimated using the TA cost, and the upper bound, given by the final MAMPD cost, we show that the decoupled approach is able to obtain near-optimal solutions in a wide range of cases.

Multi-Agent Pathfinding (MAPF) is the problem of finding paths for multiple agents such that every agent reaches its goal and the agents do not collide. Most prior work on MAPF were on grids, assumed agents' actions have uniform duration, and that time is discretized into timesteps. In this work, we propose a MAPF algorithm that do not assume any of these assumptions, is complete, and provides provably optimal solutions. This algorithm is based on a novel combination of Safe Interval Path Planning (SIPP), a continuous time single agent planning algorithms, and Conflict-Based Search (CBS). We analyze this algorithm, discuss its pros and cons, and evaluate it experimentally on several standard benchmarks.

Conflict-Based Search (CBS) is a powerful algorithmic framework for optimally solving classical multi-agent path finding (MAPF) problems, where time is discretized into the time steps. Continuous-time CBS (CCBS) is a recently proposed version of CBS that guarantees optimal solutions without the need to discretize time. However, the scalability of CCBS is limited because it does not include any known improvements of CBS. In this paper, we begin to close this gap and explore how to adapt successful CBS improvements, namely, prioritizing conflicts (PC), disjoint splitting (DS), and high-level heuristics, to the continuous time setting of CCBS. These adaptions are not trivial, and require careful handling of different types of constraints, applying a generalized version of the Safe interval path planning (SIPP) algorithm, and extending the notion of cardinal conflicts. We evaluate the effect of the suggested enhancements by running experiments both on general graphs and 2^k-neighborhood grids. CCBS with these improvements significantly outperforms vanilla CCBS, solving problems with almost twice as many agents in some cases and pushing the limits of multi-agent path finding in continuous-time domains.

Path planning in the presence of dynamic obstacles is challenging as the time dimension has to be considered. A prominent approach to tackle this problem known to be complete and optimal is the A*-based Safe-interval Path Planning (SIPP). Bounded-suboptimal variants of SIPP employing the ideas of Weighted A* (WSIPP) and Focal Search (FocalSIPP) have been introduced recently, trading-off optimality for decreased planning time. In this paper, we revisit FocalSIPP and design several secondary heuristics for Focal Search with the intention to narrow the search in the direction of a preplanned optimal single-agent path not considering dynamic obstacles. The experimental results on various maps show that the designed heuristics generally outperform the hops-to-the-goal heuristic used in the original FocalSIPP and successfully compete with WSIPP as well.

In this paper, we address the multi-agent pickup and delivery problem, a variant of multi-agent path finding.\nSpecifically, we decouple the problem into two parts: task allocation and path planning. We employ the\nany-angle safe-interval path planning algorithm introduced in our recent work and study the performance\nof several task allocation strategies. Furthermore, the proposed approach has been integrated into a control\nsystem to verify its feasibility in deployment on real robots. A key part of the system is a visual localization\nsystem which is based on the detection of unique artificial markers placed in the working environment. The\nconducted experiments show that generated plans can be safely executed on a real system.\n