MADRaS : Multi Agent Driving Simulator

Authors: Anirban Santara, Sohan Rudra, Sree Aditya Buridi, Meha Kaushik, Abhishek Naik, Bharat Kaul, Balaraman Ravindran

JAIR 2021 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section we present the results of six experiments on single and multi-agent RL for learning to drive in MADRa S. Table 4 presents a brief outline of our experiments and their individual motivations.
Researcher Affiliation Collaboration Anirban Santara EMAIL Department of Computer Science and Engineering, Indian Institute of Technology Kharagpur, Kharagpur, WB, India; Meha Kaushik EMAIL Microsoft, Vancouver, Canada
Pseudocode No The paper describes the implementation of a PID controller and mentions algorithms like Proximal Policy Optimization (PPO), but it does not include any structured pseudocode or algorithm blocks.
Open Source Code Yes MADRa S is open source2 and aims to contribute to the democratization of artificial intelligence. 2. Code available at https://github.com/madras-simulator/MADRa S
Open Datasets No The paper uses the MADRa S simulator, which is built on TORCS and inherits its assets (tracks, cars), but does not explicitly mention using or providing concrete access information for a separate, pre-existing open dataset for experimental training or evaluation.
Dataset Splits No The paper conducts experiments in a simulated environment and evaluates agents over a number of episodes (e.g., 'estimated over at least 100 episodes'), rather than using pre-defined training/test/validation splits from a static dataset.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, processor types, memory amounts) used for running its experiments.
Software Dependencies No The paper mentions using 'RLLib' for PPO implementation and 'Open AI Gym' for its interface, but does not provide specific version numbers for these or other key software dependencies.
Experiment Setup Yes Unless otherwise stated, we set the learning rate to 5e-5. The policy and value functions are modelled using fully connected neural networks with 2 hidden layers and 256 tanh units in each layer. All experiments with the track-position speed action space have a PID latency of 5 time steps. The PID parameters used for track-position speed control are given in Table 3.