In bounded-suboptimal heuristic search, the aim is to find a solution path within a given bound as quickly as possible, which is crucial when computational resources are limited. Recent research has demonstrated Weighted A* variants such as XDP that find bounded suboptimal solutions without needing to perform state re-expansions; they work by shifting where the suboptimality in the search is allowed. However, the suboptimality distribution is fixed before the search begins. This abstract describes Dynamic Suboptimality Weighted A* (DSWA*), an algorithm introduced at AAAI 2025 that allows suboptimality to be dynamically distributed at runtime based on the properties of the search.
Reinforcement learning is becoming ever more prominent in solving combinatorial search problems, in particular ones where states are images. Prior work has devised action-policy testing methodology, that identifies so-called bug states where policy performance is sub-optimal. Here we show how to leverage this methodology during the RL process, using action-policy testing to find bugs and injecting those as alternate start states for the training runs. Running experiments across six 2D games, we find that our testing-guided training often achieves similar expected reward while reducing the number of bugs.
The Hierarchical Seating Allocation Problem (HSAP) is the problem to allocate an organizational hierarchy of teams to a set of seats on a floor plan. This problem is driven by the necessity for large organizations with large hierarchies to ensure that teams with close hierarchical relationships are seated in proximity to one another, such as ensuring a research group occupies a contiguous area. Currently, this problem is managed manually leading to infrequent and suboptimal replanning efforts. To alleviate this manual process, we propose an end-to-end framework to solve the HSAP. A scalable approach to calculate the distance between any pair of seats using a probabilistic road map (PRM) and rapidly-exploring random trees (RRT) which is combined with heuristic search and dynamic programming approach to solve the HSAP using integer programming. We demonstrate our approach under different sized instances by evaluating the PRM framework and subsequent allocations both quantitatively and qualitatively.
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 of this approach is that in MAPF the agents must reach a configuration in which all agents are at their targets simultaneously. Coordinating the agents’ paths such that they occupy their targets at the same time is needlessly restrictive for LMAPF. Techniques have been proposed to indirectly mitigate this drawback. In this position paper, 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 MAPF for Lifelong (MAPF4L) and propose how to solve it by modifying several existing MAPF algorithms. A limited experimental evaluation identifies some cases where using a MAPF4L algorithm can improve the system throughput significantly.
This paper provides a survey of recent literature on the use of multi-armed bandit algorithms to solve the Boolean satisfiability problem (SAT), a well-known NP-complete problem with broad applications in academia and industry. The application of bandit algorithms in modern SAT solvers has achieved great success in recent years, as evidenced by the excellent performance of SAT solvers using bandit algorithms in SAT competitions. Bandit algorithms are classic randomized optimization algorithms that strike a balance between exploration and exploitation and can aid in designing and improving heuristics in SAT solvers. In this paper, we introduce several aspects of the application of bandit algorithms in modern SAT solvers, ranging from heuristic methods in CDCL and SLS solvers to strategies in parallel SAT solvers. The use of bandit algorithms in SAT solvers still holds great potential. In conclusion of the survey, we summarize the current issues and suggest possible future research directions.
The multi-agent pathfinding problem (MAPF) of finding conflict-free paths for multiple agents has attracted a large number of researchers in the past. The cost of the solution is commonly measured by the sum-of-costs (SOC) cost function or, less commonly, by Makespan. In this paper, we focus on the Fuel cost function, which is the number of physical steps the agents traverse. While Fuel was mentioned in many previous papers, our paper is the first to deepen into it. We introduce an A*-based algorithm and a CBS-based algorithm for Fuel. We study Fuel theoretically, showing that it can be (perhaps non-intuitively) more complex than SOC. Finally, we experimentally compare both algorithms against each other and against their SOC counter parts, studying their advantages and disadvantages.
Combinatorial optimization (CO) problems on graphs arise in various applications across diverse domains. Many of these problems are NP-hard, and heuristics have been developed to provide near-optimal solutions. In the big data era, the high dimensionality of these problems poses significant challenges for existing heuristic methods, which struggle to scale efficiently. In this paper, we propose Hierarchical DeepPruner, an adaptive framework that employs a two-stage approach to efficiently prune the search space of CO problems on graphs. Compared to state-of-the-art pruning heuristics, our algorithm offers two key advantages: 1) it does not require extensive feature engineering or domain-specific knowledge, and 2) it outperforms all previous methods while consistently pruning over 95% of the ground set, resulting in up to several of tenfold speedups—typically with minimal impact on solution quality. Additionally, our algorithm can successfully reduce the search space of instances even if they lie outside the training distribution, resulting in small optimality gaps across multiple budgets
Multi-Agent Path Finding (MAPF) seeks collision-free paths for multiple agents from their respective starting locations to their respective goal locations while minimizing path costs. Although many MAPF algorithms were developed, most of them rely on a common assumption on synchronized actions, where the actions of all agents start at the same time and always take a time unit. This assumption may limit use of MAPF planners in practice. To get rid of this assumption, recently, an algorithm called Loosely Synchronized Rule-Based Planning (LSRP) is proposed, which can find sub-optimal solutions for many agents. However, LSRP often finds poor quality solutions due to its unbounded sub-optimality. This paper develops a new anytime planner called LSRP* that can keep improving solution quality after the initial solution is obtained until the runtime budget depletes. We analyze the properties of LSPR* and test it against several baselines with up to 1000 agents in various maps. LSRP* can handle up to 25% more agents than LSRP and can reduce up to 40% of the solution cost found by LSRP.
We introduce a priority queue data structure, called a bucket heap, which generalizes the bucket queue commonly used to accelerate A* search for shortest-path problems with a small range of integer transition costs. Unlike a bucket queue, a bucket heap speeds up priority queue operations for bounded-suboptimal and Anytime A* algorithms guided by non-admissible node evaluation functions. It also provides direct access---without any additional overhead---to the underlying bucket queue of A*, which we show can be used to improve search performance in further ways.
This work builds upon existing task and motion planning (TAMP) frameworks by integrating pre-trained Sequencing Task-Agnostic Policies (STAP) and Effort Level Search (ELS) to create a hierarchical approach that decouples high-level task decisions from low-level motion execution. The method enhances the planning process by incorporating a novel success rate estimator (P ), which provides more accurate task success predictions than traditional Q-value estimators. We formalize the problem of long-horizon manipulation tasks, where high-level decisions are made in discrete spaces and low-level actions are executed in continuous space. To guide the search process efficiently, we leverage the infinite completion tree structure of ELS, which dynamically adjusts computational resources based on task complexity. Empirical results demonstrate that our approach significantly improves planning efficiency and execution reliability, outperforming traditional methods by reducing the search space and computational overhead. Our work highlights the effectiveness of combining learned skills from STAP with ELS and P in a hierarchical structure, laying the foundation for scalable robotic planning in complex, real-world manipulation tasks.
This paper investigates a Traveling Salesman Problem with Time Windows and Vacant Penalties (TSP-TW-VP), which plans a path to service a set of machines at different locations within their respective time windows while minimizing two objective functions: the finish time and penalty for machine vacancy. There is often no single solution that optimizes both objectives simultaneously, and the problem thus seeks the Pareto-optimal solutions. TSP-TW-VP generalizes TSP-TW and is therefore NP-hard. To solve the problem, this paper develops an algorithm called Search with Look-Ahead Pruning (S-LAP) that is guaranteed to find all Pareto-optimal solutions for TSP-TW-VP. S-LAP gains computational efficiency by introducing a novel look-ahead pruning rule, and a fast dominance checking method based on both the objective functions and path history. Experimental results show that the proposed look-ahead pruning and fast dominance can speed up the search for 2-8 times over 4 different datasets.
This extended abstract presents new empirical results of recently introduced Critical Section Macro-operators (CSMs) whose design is inspired by using lockable resources in critical sections in parallel computing. In particular, we provide results on the IPC-2023 learning track domains and four planners, including the winner of the agile track of the IPC-2023 and a lifted planner.
In the context of pathfinding, first-order logic allows for the expressive specification of goals. Using negation as failure, one can specify what must not be true in a goal state instead of what must be true, which can result in succinct goal specifications while also being computationally advantageous. However, due to non-monotonicity, integration of negation as failure can be cumbersome. To address this problem, we introduce conflict-driven goal reaching (CDGR), a conflict-driven algorithm for reaching goals specified with non-monotonic logic that refines a search for a goal state based on conflicts encountered during search. Our results show that CDGR results in significantly shorter paths and can significantly speed up search when compared to not taking conflicts into consideration. Furthermore, our results show that finding paths to goals can be much more efficient when goals are specified with negation as failure instead of without negation as failure.
We consider a game that was played in a German television show that is similar to the sorting balls puzzle. In it, we are assumed to move one colored ball after another in a set of colored tubes so that in the end, each ball is in the tube of its color. We are allowed to use one additional (uncolored) tube. We show general properties for solvability and that the problem of minimizing the number of moves is NP-hard, which is done by a reduction from the Feedback Arc Set Problem. Furthermore, we give an implementation of an algorithm to compute such a minimal sequence of moves. The algorithm is based on breadth-first search and accelerated by a lower bound on the number of moves from the current configuration to the final one that is obtained by solving a small instance of the Feedback Arc Set Problem. Our experiments show that instances with 7 colored tubes of height 4 can be solved in a reasonable amount of time and that the number of tubes is much more critical for the running time than the heights of the tubes.
Tree Cache is a lightweight pre-processing approach to grid path finding which works by generating a shortest path tree: from a root cell to all cells in the map. During online search Tree Cache simply follows the tree: from start and target towards the root, stopping at the first common cell. Although Tree Cache is fast, the resulting paths have no solution quality guarantees. In this paper we improve Tree Cache, in terms of speed and solution quality, by combining symmetry breaking ideas from Jump Point Search. Our new algorithm, Jump Spanning Tree Search (JSTS), can usually generate paths with low average sub-optimality in under one microsecond -- up to two orders of magnitude faster than Tree Cache. We then extend JSTS to derive a new and very fast bounded suboptimal search, which guarantees solution quality in single-digit microseconds on average. Our results establish a remarkable new level of performance in the area. In particular, we show JSTS approaches and often improves upon the output complexity of an idealised oracle, which simply reads off and returns a corresponding but optimal solution path.
Enhancing Greedy Best First Search (GBFS) with stochastic exploration will often greatly improve search performance. In this work, we show that one way exploration does so is by helping the search find states that are "easy" for standard GBFS without exploration. In particular, we show that in problems in which standard GBFS struggles and exploration helps, there are often many states that are reachable from the initial state that standard GBFS can quickly find solutions from. Many such states are actually outside the Bench Transition System (BTS) --- which is a structure that contains all states that standard GBFS may encounter --- meaning GBFS cannot reach them without using exploration. To allow exploration mechanisms to better exploit the existence of such states, we introduce a method called locally greedy probes. Upon a successor having an improved heuristic from its parent, locally greedy probes pause exploration and greedily hill-climb along a single path as long as heuristic improvements keep occurring. Our empirical evaluation shows that this approach is effective at enhancing several exploration mechanisms in a variety of classical planning domains.
