# ARISE-Initiative/robosuite

robosuite: A Modular Simulation Framework and Benchmark for Robot Learning

Repository: https://github.com/ARISE-Initiative/robosuite
Canonical: https://ross.abutalabs.com/products/robosuite
Homepage: https://robosuite.ai
Language: Python
License: NOASSERTION
License Family: other
Topics: robotics, robot-manipulation, reinforcement-learning, physics-simulation, robot-learning
Last push: 2026-07-11T23:47:48+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 92, release rhythm 30, longevity 100
- inputs: {"age_days": 2869, "days_push": 53, "days_rel": 252, "gap_med": 210.0, "n_releases_24m": 3}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2581, forks 760 (observed 2026-08-28T04:07:02.429838+00:00)

## What it is
robosuite is a modular simulation framework powered by the MuJoCo physics engine for robot learning, offering standardized benchmark environments for reproducible research. It supports diverse robot embodiments including humanoids, custom robot composition, composite controllers, teleoperation devices, and photo-realistic rendering.

## Use cases
- train reinforcement learning agents for robot manipulation tasks
- benchmark imitation learning algorithms on standardized robot tasks
- simulate humanoid and multi-robot embodiments in MuJoCo
- design custom robot simulation environments with modular components
- collect teleoperated demonstration data for robot learning
- run reproducible robotics research without physical hardware
- apply domain randomization and dynamics randomization for sim-to-real transfer

## When to choose
- you need standardized benchmark environments for robot manipulation research
- you want to train RL or imitation learning policies in simulation before deploying to hardware
- you need support for diverse robot embodiments including humanoids
- you require reproducible, community-adopted simulation for embodied AI research

## When to avoid
- you need a physics engine itself rather than a task framework built on one
- your focus is mobile robot navigation or locomotion rather than manipulation
- you need photorealistic simulation beyond what MuJoCo rendering provides
- you require a non-Python or real-time hardware control stack

## Facets
- artifact type: framework
- maturity: active
- function: simulation, machine-learning, reinforcement-learning, benchmarking, robotics
- domain: robotics, reinforcement-learning, simulation, machine-learning
- platform: python, windows
- tags: mujoco, robot-learning, robot-manipulation, imitation-learning, benchmark-environments, sim-to-real, embodied-ai, research, linux, macos

## Member repositories
- ARISE-Initiative/robosuite (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:02.429838+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:22:36.730958+00:00, confidence not recorded.
  - readme: https://github.com/ARISE-Initiative/robosuite (fetched 2026-08-28T04:07:02.429838+00:00, sha 2ce33e703a18)
  - homepage: https://robosuite.ai (fetched 2026-08-29T10:05:09.994224+00:00, sha cac2bb69860e)
  - site_page: https://robosuite.ai/docs/overview.html (fetched 2026-08-29T10:05:10.003919+00:00, sha 0c5d94f92074)
  - site_page: http://robosuite.ai/docs/overview.html (fetched 2026-08-29T10:05:10.006502+00:00, sha 9246f05103f3)
  - registry_pypi: https://pypi.org/pypi/robosuite/json (fetched 2026-08-29T10:05:10.012087+00:00, sha 6e25a14434a9)
- Data as of 2026-08-30T08:39:29.467469+00:00.