Where the theory must survive reality.
Robotics and aerospace are current validation directions. They are research tracks, not product verticals.
Manipulator and mobile-robot control after resource loss.
Current implementation work maps naturally to robotic platforms because sensing, actuation, compute and communication resources can be isolated and degraded independently.
| Failure | Residual question | Væringjar II mechanism |
|---|---|---|
| actuator loss | what motion/control authority remains? | \(\boldsymbol{\rho}(t)\mapsto\mathcal F_{\mathrm{feas}}(\boldsymbol{\rho}(t))\mapsto c^\star(t)\) |
| sensor degradation | what state can still be estimated? | HealthMonitor + estimator |
| compute / communication loss | which control path remains executable? | service/resource remapping |
| multiple faults | is continued operation still stable? | configuration search + stability gate |
Flight-control problems with residual control authority.
The aerospace track starts from unmanned and flight-control cases where the damaged plant cannot be repaired in mission, but a reduced set of actuators, sensors, power or communication resources may still support a bounded objective.
| Failure | Residual question | Research target |
|---|---|---|
| partial actuation | is the damaged vehicle still controllable? | reduced plant model + control reformulation |
| sensor loss | what navigation/control state remains observable? | health estimation + degraded state estimation |
| power degradation | which functions remain mission-feasible? | resource/function prioritisation |
| communication degradation | what autonomy must remain local? | configuration / service reallocation |
A useful Væringjar II case has four properties.
Irreversible change
The object is physically or structurally different after the failure.
Residual resources
Some control authority remains; the system is not simply dead.
Mission value
There is value in preserving some function instead of immediately stopping.
Verifiable boundary
Continued operation can be tested against stability, timing and safety constraints.
Have a system that can challenge the architecture?
We are looking for research laboratories, engineering teams and experimental platforms—not customers—to help test where the approach works and where it fails.