# MIT-SPARK/Hydra

A system for building 3D Scene Graphs from sensor data in real-time

Repository: https://github.com/MIT-SPARK/Hydra
Canonical: https://ross.abutalabs.com/products/mit-spark-hydra
Language: C++
License: BSD-2-Clause
License Family: permissive
Last push: 2026-08-31T22:55:48+00:00

## Health v2 (maintenance only)
Score: 68/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 8, longevity 100
- inputs: {"age_days": 1581, "days_push": 2, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1142, forks 144 (observed 2026-09-01T02:13:55.548293+00:00)

## What it is
Hydra is a C++ system that incrementally builds hierarchical 3D Scene Graphs from sensor data in real time. It is developed by MIT SPARK as research software for robotic spatial perception, with ROS2 integration and Python bindings.

## Use cases
- build a 3D scene graph from robot sensor data in real time
- incrementally construct hierarchical spatial maps for a mobile robot
- run open-set semantic segmentation and fuse it into a scene graph
- evaluate spatial perception systems on robotics datasets
- provide spatial scene representations for downstream robot planning or VLM tasks

## When to choose
- you need real-time hierarchical 3D scene graph construction from sensor streams
- you are doing robotics research on spatial perception or semantic mapping
- you work in a ROS2 environment and want an open-source scene graph pipeline

## When to avoid
- you need a simple standalone SLAM or mapping tool without scene graph structure
- you require ROS1 support, which has been archived
- you need a production-tested commercial-grade perception stack

## Facets
- artifact type: library
- maturity: active
- function: computer-vision, simulation, machine-learning
- domain: robotics, computer-vision, artificial-intelligence
- platform: cpp
- tags: 3d-scene-graph, spatial-perception, slam, real-time, ros2, robotics-perception, linux, ros

## Member repositories
- MIT-SPARK/Hydra (main) score 68

## Provenance
- Observed fields: from GitHub, fetched 2026-09-01T02:13:55.548293+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 2026-08-30T06:36:50.018686+00:00, confidence not recorded.
  - readme: https://github.com/MIT-SPARK/Hydra (fetched 2026-09-01T02:13:55.548293+00:00, sha 68945fe9f6d5)
- Data as of 2026-08-30T08:39:29.467469+00:00.