Toward Estimating Others' Transition Models Under Occlusion for Multi-Robot IRL

Authors: Kenneth Bogert, Prashant Doshi

IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate m IRL /T +Int in an application domain introduced by Bogert and Doshi [2014] involving mobile robots, I and J, patrolling hallways using cyclical trajectories as shown in Fig. 2. ... We report the success rates of the physical runs in Table 1.
Researcher Affiliation Academia Kenneth Bogert and Prashant Doshi THINC Lab, Department of Computer Science University of Georgia, Athens, GA 30602 {kbogert,pdoshi}@uga.edu
Pseudocode No The paper describes mathematical formulations and iterative procedures in prose but does not provide a clearly labeled pseudocode or algorithm block.
Open Source Code No The paper does not provide any explicit statement about releasing the source code for its methodology or a link to a code repository.
Open Datasets No The paper describes data generated from simulations and physical robot experiments but does not provide concrete access information (link, DOI, repository, or formal citation) for a publicly available or open dataset.
Dataset Splits No The paper evaluates performance through simulations and physical runs but does not provide specific dataset split information (percentages, sample counts, or detailed splitting methodology) for training, validation, and testing.
Hardware Specification Yes Each robot in our simulations and physical experiments is a Turtle Bot equipped with a Kinect, which provides a camera and an infrared ranging sensor. The bases include an i Robot Create or a Kobuki.
Software Dependencies Yes Each robot also has a laptop running ROS Hydro on Ubuntu 12.04. A robot identifies another by detecting its unique color signature using CMVision s blob finder. ROS s default actuator and sensor models for the Turtle Bot and the default local motion planner in move base are used for navigation. Each robot localizes in a predefined map using the adaptive Monte Carlo localization available in ROS. The virtual simulations are performed in Stage.
Experiment Setup Yes The learner s transition function models the probability of any of its own action failing at 2.5%. ... We experiment with m IRL /T +Int, m IRL +Int fixing a transition success rate of 0.9, and Random in both scenarios: when J s left wheel is artificially damaged thereby slowing it down and creating an uneven trajectory, and when the patrollers are operating properly.