Technology

Tianshu — City-Scale Robot Command Hub

A live scene graph at the centre, connecting robots, cameras, sensors, GIS maps and business systems for sensor fusion, incident assessment, task orchestration, multi-robot coordination and an operating loop.

The closed loop

  1. Sensor intake
  2. live scene graph
  3. incident assessment
  4. task orchestration
  5. coordinated execution
  6. operating loop

Platform modules

  1. Sensor intake and data fusion

    • Robots, cameras, radar and IoT across multiple sources
    • Traffic, campus and fire-safety business data fused
    • GIS maps, device state and live situational data
  2. Live scene graph

    • People, vehicles, objects, devices and events modelled uniformly
    • Position, state, risk and task updated continuously
    • One shared picture of the public-safety situation
  3. Incident assessment and risk reasoning

    • Anomalous behaviour and incident detection
    • Spatio-temporal correlation and risk grading
    • Response recommendations and alert triggers
  4. Task orchestration and dispatch

    • Task decomposition and prioritisation
    • Allocation by route, zone and robot
    • Multi-robot coordination, handover and dynamic replanning
  5. Coordinated execution and safety control

    • Robots, PTZ cameras, broadcast and lighting act together
    • Human-in-the-loop command and remote takeover
    • Permissions, audit trail and safety fallback
  6. Operating loop and data assets

    • Execution results returned, incidents reviewed, metrics dashboarded
    • Failure cases retained as long-tail scenario assets
    • Fed back into Tianzhi training for policy iteration

See everything, judge correctly, dispatch effectively, stay accountable, keep learning — turning scattered devices into a coordinated, dispatchable, evolving public-safety network.