Assistive Gym is a physics-based simulation framework for physical human-robot interaction and robotic assistance. It targets researchers and engineers in robotics and reinforcement learning, enabling the use of existing RL algorithms to train robots for human-robot collaboration tasks.

How It Works

Assistive Gym integrates with the OpenAI Gym interface, providing a standardized API for defining and interacting with simulated human-robot scenarios. It leverages a physics engine (likely PyBullet, given the syntax examples) to model realistic human and robot dynamics, including human joint limits and preferences. This physics-based approach allows for the training of robots that can safely and effectively assist humans in tasks like feeding or manipulation.

Quick Start & Requirements

• Install: pip3 install git+https://github.com/Healthcare-Robotics/assistive-gym.git
• Recommended Python: 3.6 or 3.7.
• Google Colab: Fully supported, no local installation required. See Wiki-Google Colab.
• Full Install: Requires cloning the repo and installing in a virtual environment (pip3 install -e .).
• Getting Started Guide: 10 Minute Getting Started Guide

Highlighted Details

• Supports four collaborative robots: PR2, Jaco, Baxter, and Sawyer.
• Includes customizable human models with 40 actuated joints and realistic pose-dependent joint limits.
• Offers mobile base control with dynamic ground friction and slip for domain randomization.
• Integrates with iGibson for visually realistic home environments and supports SMPL-X human mesh models.
• Provides pretrained control policies and tools for training new policies.

Maintenance & Community

• Version 1.0 released, with significant updates including Google Colab support and multi-robot control.
• Active development indicated by recent feature additions.
• Links to documentation and examples are provided within the README.

Licensing & Compatibility

• The README does not explicitly state a license. However, the project is hosted on GitHub, suggesting a common open-source license. Further clarification on licensing is recommended for commercial use.

Limitations & Caveats

• The recommended Python versions (3.6/3.7) are older, potentially requiring environment management for users on newer Python versions.
• The license is not explicitly stated, which could be a concern for commercial adoption.