Effective Integration of Weighted Cost-to-Go and Conflict Heuristic within Suboptimal CBS

Authors: Rishi Veerapaneni, Tushar Kusnur, Maxim Likhachev

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental Results We test our methods with different numbers of agents, in increments of 50, on 8 diverse maps from Stern et al. (2019) and report the mean values across 5 seeds.
Researcher Affiliation Academia Rishi Veerapaneni, Tushar Kusnur, Maxim Likhachev Robotics Institute, Carnegie Mellon University {rveerapa, tkusnur, mlikhach}@andrew.cmu.edu
Pseudocode Yes Algorithm 1: Suboptimal CBS low level focal search planner Input: nstart, at Goal(), Paths PI of other agents Output: Lower bound LB on optimal path cost, Path from nstart with sub-optimality wso (i.e. cost wso LB)
Open Source Code No The paper mentions a detailed version with supplementary material at an arXiv link (https://arxiv.org/abs/2205.11624), but does not explicitly state that its own source code is provided there or elsewhere.
Open Datasets Yes We test our methods with different numbers of agents, in increments of 50, on 8 diverse maps from Stern et al. (2019)
Dataset Splits No The paper tests on '8 diverse maps' and reports 'mean values across 5 seeds' but does not specify explicit training/validation/test dataset splits or cross-validation methodology.
Hardware Specification No The paper states 'The speed up Smethod = Tbaseline/Tmethod is reported to normalize differences in hardware' but does not provide any specific hardware details such as CPU, GPU models, or memory.
Software Dependencies No The paper refers to using EECBS (Li, Ruml, and Koenig 2021) and its open-source codebase, but it does not specify any particular software dependencies with version numbers (e.g., Python version, library versions) used for its own implementation or experiments.
Experiment Setup Yes We use wso = 2 and a timeout of 300 seconds in all our experiments unless otherwise specified. In all figures, if a method failed (timed out on all 5 seeds) on a particle number of agents on a map, we do not report larger number of agents.