KORTHOS / ECOSYSTEM / PRODUCT

Galaxea R1 Lite

About

About Galaxea R1 Lite

Galaxea R1 Lite is a dual-arm mobile platform for embodied-AI research, data collection and application development. Two six-axis A1X arms and G1 force-controlled parallel grippers sit above a movable torso and omnidirectional chassis. The torso changes working height, while the base supports movement between handling locations. Each arm is rated for three kilograms, with a five-kilogram maximum.

The standard package includes two monocular wrist cameras, controllers and calibration hardware. Additional depth and fisheye cameras and the R1 Lite-T teleoperation device are optional; dexterous hands and VR equipment are supplied separately. An Intel Core i9 computer handles onboard processing. Galaxea offers the modular system through business and education enquiries, allowing teams to select sensing and control equipment around their data-collection or manipulation experiments.

25 FEB 2025

Galaxea launches R1 Series for wheeled humanoid manipulation

Capabilities

Source-backed capabilities

  • Operating environments: AtomVLA physical tabletop evaluation, CORAL held-out physical scenes and task stations, EvoScene-VLA indoor-cleaning research station
  • Control modes: Autonomous
  • Atom stack bowls

    Research demonstration - 2026-03-09 - Pick up a bowl and stack it on another bowl. - environment: AtomVLA physical tabletop evaluation

    Source
  • Atom fruit basket

    Research demonstration - 2026-03-09 - Pick up fruit and place it in a basket. - environment: AtomVLA physical tabletop evaluation

    Source
  • Atom hang cup

    Research demonstration - 2026-03-09 - Hang a cup on a mug rack. - environment: AtomVLA physical tabletop evaluation

    Source
  • Atom open drawer

    Research demonstration - 2026-03-09 - Open a closed drawer. - environment: AtomVLA physical tabletop evaluation

    Source
  • Atom fold shirt

    Research demonstration - 2026-03-09 - Fold a T-shirt into the target configuration. - environment: AtomVLA physical tabletop evaluation

    Source
  • Atom fold towel

    Research demonstration - 2026-03-09 - Fold a towel twice into the target configuration. - environment: AtomVLA physical tabletop evaluation

    Source
  • Coral close book

    Research demonstration - Use both hands to close an open book until its covers touch. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral press tissue

    Research demonstration - Pick up a tissue, place it into a tissue box and press it down. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral uncap pen

    Research demonstration - Grasp a pen and use both hands to separate its cap from the body. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral empty food

    Research demonstration - Hold a bowl and empty its contents into a trash bin without dropping the bowl. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral arrange bottles

    Research demonstration - Move a selected water bottle beside another without knocking the others over. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral fold shirt

    Research demonstration - Use both hands to fold a black T-shirt into the target shape. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral insert straw

    Research demonstration - Use both hands to insert a straw into a milk carton opening. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral discard bottle

    Research demonstration - Pick a plastic bottle from the floor and place it fully inside a trash bin. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral open doors

    Research demonstration - Grasp a door handle and push the door open far enough for passage. - environment: CORAL held-out physical scenes and task stations

    Source
  • Coral elevator buttons

    Research demonstration - Press a selected elevator button until it lights, without activating adjacent buttons. - environment: CORAL held-out physical scenes and task stations

    Source
  • Evoscene mirror

    Research demonstration - Wipe ketchup stains from a mirror. - environment: EvoScene-VLA indoor-cleaning research station

    Source
  • Evoscene sink

    Research demonstration - Acquire a cloth and wipe ketchup stains from a sink basin. - environment: EvoScene-VLA indoor-cleaning research station

    Source
  • Evoscene cutting board

    Research demonstration - Wipe ketchup stains from a cutting board. - environment: EvoScene-VLA indoor-cleaning research station

    Source

Operations

  • Shipments

    began shipping - 2025-08-25 - R1 customers - humanoid/semi-humanoid robotics - Forbes reported that Galaxea had begun shipping R1 humanoid robots. Forward-looking unit targets from the article are not counted here.

    Source

Supply

Supply / Components

  • Galaxea R1 Lite

    Jetson Agx Orin 32gb For R1 Lite - Compute & Control - NVIDIA - Galaxea R1 Lite integrates an NVIDIA Jetson AGX Orin 32GB compute module - source: docs.galaxea-dynamics.com

    Source

Research

  • GigaBrain-0.7: Scaling Embodied Foundation Models to Emergent Capabilities with a Three-System Architecture

    2026-08-16 - Research Publication - GigaBrain-0.7 combines action control, understanding and planning, and world-model prediction in a three-system embodied model. GigaAI reports training on over 37,000 hours of heterogeneous data and physical evaluation on Maker H01 and AgileX PiPER-family manipulators.

    Source
  • EvoScene-VLA: Evolving Scene Beliefs Inside the Action Decoder for Chunked Robot Control

    2026-05-21 - Research Publication - The authors argue for a persistent action-updated scene state across control calls, and introduce EvoScene-VLA.

    Source
  • CORAL: Scalable Multi-Task Robot Learning via LoRA Experts

    2026-03-10 - Research Publication - To address this dilemma, the authors present CORAL, a backbone- and embodiment-agnostic framework designed primarily to mitigate multi-task interference while remaining naturally extensible to a continuous stream of new tasks.

    Source
  • AtomVLA: Scalable Post-Training for Robotic Manipulation via Predictive Latent World Models

    2026-03-09 - Research Publication - To tackle this problem, the authors propose \method, the first subtask-aware VLA framework integrated with a scalable offline post-training pipeline.

    Source
  • Galaxea publishes Open-World Dataset and G0 dual-system VLA model for R1

    2025-08-30 - Research Publication - We present Galaxea Open-World Dataset, a large-scale, diverse collection of robot behaviors recorded in authentic human living and working environments.

    Source
  • FiveAges robot_usds simulation asset library

    Research Publication - A shared library of robot USD assets for ROS 2 control and Isaac Sim development, with configurable robot and end-effector models.

    Source
  • GalaxeaVLA: G0.5 vision-language-action model and robot workflows

    Research Publication - Galaxea's VLA software release provides model fine-tuning, simulation evaluation and real-robot deployment workflows, including R1 Pro configurations.

    Source
  • GalaxeaDP: R1 diffusion-policy simulation workflows

    Research Publication - GalaxeaDP provides diffusion-policy training, generated-data conversion and simulation evaluation for Galaxea R1 robots. Its source includes R1 Pro and R1 Lite task configurations and distinguishes current simulation workflows from planned real-world support.

    Source
  • Galaxea Manipulation Simulator

    Research Publication - Galaxea Manipulation Simulator provides reusable manipulation benchmarks, expert demonstration generation and baseline training/evaluation workflows for R1, R1 Pro and R1 Lite. It includes more than 30 environments across 17 tasks and exact robot/controller registrations.

    Source
  • Galaxea Diffusion Policy implementation

    Research Publication - GalaxeaDP implements diffusion-policy training, open-loop evaluation and simulated deployment with GalaxeaManipSim and LeRobot data formats. Named task configurations distinguish R1 Pro end-effector/joint policies and R1 Lite joint policies.

    Source

Timeline

  • Galaxea presents its Open-World Dataset and G0 model

    2025-08-30 - Research Publication - The dataset comprised 500 hours recorded on Galaxea R1 Lite across more than 150 tasks and 50 scenes, with subtask-level language annotations. G0 combined vision-language planning with vision-language-action execution, trained through cross-embodiment pretraining, single-embodiment pretraining and task-specific adaptation.

    Source

Context

Market context

Galaxea R1 is a mobile manipulation robot for embodied AI development and indoor task automation.

  • Target environment: Robotics labs, developer spaces and structured indoor work environments.
  • Workflow context: Mobile manipulation, object handling, robot learning, indoor navigation and task execution.
  • Deployment model: Mobile manipulator used for research, development and controlled task demonstrations.
  • Commercial maturity: Public robot product from Galaxea Dynamics.
  • Adoption constraints: Requires software integration, task setup, manipulation reliability and controlled operating conditions.
  • Market position: Mobile manipulation platform for teams developing embodied AI and practical indoor robot behaviours.
  • Adjacent products: Indoor mobile manipulator for robot learning, object handling and embodied AI workflows.

Comparable Systems

Related Systems