Track D
Physical
How do Route, Ground and Assure apply where errors have physical consequences?
The same Route, Ground, Assure stack where errors have physical consequences.
Why it matters
Physical trials are planned. Claims will name the testbed stage and report sim-to-real gaps.
Directions
1. D1: Swarm coordination under degraded conditions.
D1: Swarm coordination under degraded conditions.
2. D2: Assurance for learned autonomy (runtime monitors and safety shields).
D2: Assurance for learned autonomy (runtime monitors and safety shields).
3. D3: Language and agents as robot interfaces, with verification gates.
D3: Language and agents as robot interfaces, with verification gates.
4. D4: Assistive robotics with demonstrable safety.
D4: Assistive robotics with demonstrable safety.
This track is described at the program level only. Named projects, hardware catalogues and market claims are withheld until positioning is approved.
Evidence rule: every claim names its testbed stage (simulation, hardware-in-the-loop, controlled indoor, outdoor field) and reports sim-to-real gaps.
Planned paths
- D1: Swarm coordination under degraded conditions
- D2: Assurance for learned autonomy (runtime monitors and safety shields)
- D3: Language and agents as robot interfaces, with verification gates
- D4: Assistive robotics with demonstrable safety
Testbed ladder
Four planned stages: simulation, hardware-in-the-loop, controlled indoor, outdoor field. Status comes from the research data file.
1
Simulation
ROS 2, Gazebo
Planned2
Hardware-in-the-loop
Flight controller or board with simulated environment
Planned3
Controlled indoor trials
Small, safe, repeatable
Planned4
Outdoor field trials
Only with permissions and a safety pilot
Planned
Status: Planned (pattern), In progress (accent border), Done (filled).
View as table
| Step | Tools | Status |
|---|---|---|
| Simulation | ROS 2, Gazebo | Planned |
| Hardware-in-the-loop | Flight controller or board with simulated environment | Planned |
| Controlled indoor trials | Small, safe, repeatable | Planned |
| Outdoor field trials | Only with permissions and a safety pilot | Planned |
Digital stack versus physical stack
Route, Ground and Assure compared in the digital version and the physical version.
| Layer | Digital | Physical |
|---|---|---|
| Route | Policy-driven routing across models and providers | Routing of planners, agents and comms under degraded conditions |
| Ground | Provenance-first knowledge graph of what the system knows | World state and sensor evidence with named sources |
| Assure | Quality contracts before an output ships | Runtime monitors and safety shields before motion |
View as table
| Layer | Digital | Physical |
|---|---|---|
| Route | Policy-driven routing across models and providers | Routing of planners, agents and comms under degraded conditions |
| Ground | Provenance-first knowledge graph of what the system knows | World state and sensor evidence with named sources |
| Assure | Quality contracts before an output ships | Runtime monitors and safety shields before motion |
Planned metrics
Values show Pending until measurements exist.
Mission completion under injected faults
Pending
Share of missions completed after named faults are injected
Time to re-plan after losing an agent
Pending
Time from agent loss to a revised feasible plan
Unsafe-plan interception rate
Pending
Share of unsafe plans stopped by the monitor before execution
False-block rate
Pending
Share of safe plans blocked by the monitor
Connects to
The lab designs and studies five research objects. They are planned artifacts on the agenda, not products with published measurements.
Work with HOPN Lab
Enterprise buyers, academic collaborators, students and funders in Europe can write to us.