Skip to main content

Enable the RoboXStudio Platform

Key launch checklist (must-read)

Follow this sequence to avoid data issues:

  1. Request platform access: Get a RoboXStudio account and set permissions (admin/developer/collector).
  2. Deploy RoboDriver-Server: Complete server deployment and verify connectivity with RoboXStudio.
  3. Start robot runtime: Bring up the robot environment (ROS1/ROS2 nodes, Dora services, or vendor SDK drivers) and verify the basic data path.
  4. Start RoboDriver adapters: Run the robot-specific robodriver_robot_* (device adapter) and robodriver_teleoperator_* (teleoperation) modules to ensure data is received.
  5. Dispatch and execute tasks: Configure and publish capture tasks in RoboXStudio; RoboDriver-Server forwards them to RoboDriver for execution.

System architecture

The embodied intelligence platform around RoboDriver has three core components with clear responsibilities:

ComponentCore responsibilitiesInteraction
RoboDriverLocal robot control, capture task execution, real-time data capture, local playbackReceives tasks/commands from RoboDriver-Server and reports status/raw data
RoboDriver-ServerTask/command forwarding, platform command handling, data ingestion/encoding/upload, unified status managementReceives tasks from RoboXStudio, forwards to the target RoboDriver; uploads captured data to storage
RoboXStudioVisual task design, task publishing/distribution, progress monitoring, role-based permissionsSends tasks to RoboDriver-Server and displays system status and data outputs

RoboDriver platform architecture

Core workflow

  1. Design capture tasks in RoboXStudio (frequency, data types, motion sequences) and assign to operators/robot clusters.
  2. RoboDriver-Server receives the task and forwards capture commands to the target RoboDriver.
  3. RoboDriver executes the task, capturing sensor/motion data and storing it in the CoRobot format.
  4. After completion, RoboDriver-Server handles encoding/format conversion and uploads to cloud or designated storage.
  5. Monitor progress in RoboXStudio or replay locally via RoboDriver for verification.

RoboXStudio usage

Apply for access

https://ei2data.baai.ac.cn/home

Core capabilities

  • Visual task design: configure robot type, capture scenarios, and data labels.
  • Task publishing and distribution: assign tasks by role.
  • End-to-end monitoring: track execution progress and uploads in real time.
  • Data management: preview/filter/export captured data; link with labeling tasks.
  • Permission management: role-based control for admins/developers/collectors.

Documentation

Full manuals, permission setup, and task design guides: RoboXStudio docs

RoboDriver-Server deployment

Quick start

RoboDriver-Server repository

Core features

  • Forward capture tasks/commands and sync status.
  • Receive platform tasks/commands from RoboXStudio.
  • Ingest data from RoboDriver, handle encoding and standardization.
  • Batch/real-time uploads to cloud object storage or local file systems.
  • Unified management and reporting of robot online status and capture progress.

Documentation

See: RoboDriver-Server docs

Notes

  • After starting the robot runtime, verify commands like rostopic list (ROS) / dora status (Dora) return data to ensure connectivity.
  • After starting RoboDriver adapters, run a local test (e.g., preview capture) before accepting tasks from the platform.
  • If tasks do not arrive or uploads fail, first check network connectivity and that components were started in the correct order.