Remote Maintenance and Teleoperation

Activated components demand robotic and teleoperated intervention; L7 coordinates remote handling against the twin so humans stay out of high-radiation zones.

Why maintenance must be remote

After operation, the breeder's first wall, blanket, and divertor are activated by 14 MeV neutron exposure, and even the burner, with its lower 5.44% neutron fraction, activates nearby structure. Hands-on maintenance in these zones is limited by dose. L7's remote-maintenance service coordinates robotic and teleoperated intervention so most in-vessel and near-vessel work is done without a human entering the radiation field.

Twin-guided teleoperation

Teleoperation is not blind driving; it is guided by the digital twin. The twin provides the remote operator or autonomous manipulator with a live geometric and radiation model of the workspace: where activated hotspots are, where residual field remains, and where the component under service sits relative to tooling. Motion plans are checked against this model before execution so a manipulator does not collide with structure or linger in a hotspot.

# validate a remote-handling motion plan against the twin
def safe_plan(plan, comp):
    twin_scene = twin_geometry(comp.location)      # live geometry
    dose_field = twin_dose_map(comp.location)       # residual activation
    if collides(plan, twin_scene): return reject('collision')
    if cumulative_dose(plan, dose_field) > tool_limit:
        return reject('dose to tooling/electronics')
    return approve(plan)                            # then execute teleop

Radiation degrades the robots too: manipulator electronics and sensors have dose limits, so L7 tracks cumulative dose to the tooling as a resource, scheduling tool rotation and shielding just as it schedules component maintenance. A remote-handling campaign is planned to keep both the components and the tools within limits.

Latency and autonomy are traded deliberately. Fine manipulation may be teleoperated with a human in the loop; repetitive or well-modeled tasks run autonomously with the twin as supervisor and the operator as approver. Either way, the same envelope-and-collision checks apply.

Remote maintenance is developed today against the twin's geometry and simulated activation fields. The procedures are validated in simulation and mockup before FOAK, so that when real activated hardware exists from ~2030, intervention is rehearsed rather than improvised.