R / 05 / Active programme
Intelligence becomes accountable when it can move a machine.
Diyar connects perception, real-time control, simulation, embedded execution, and edge policy so autonomous behaviour can remain bounded when networks, compute, and sensor confidence change.
Research question
How should an autonomous machine preserve safe authority when sensors, networks, compute, and the environment become uncertain?
Workstreams
The programme is worked as a system.
Perception & state
Connect sensing, estimation, confidence, and world state to decisions that remain observable and testable.
Control & real time
Engineer deterministic control paths, timing budgets, fallback behaviour, and hardware-aware constraints.
Simulation & validation
Use simulation, replay, hardware-in-the-loop, and scenario coverage to move evidence toward the physical system.
Edge policy & fleet
Keep local authority explicit while signed policy, telemetry, updates, and fleet evidence cross the network boundary.
Evidence gate
What would move the claim forward?
Research maturity advances through named evidence, not through the passage of time or the confidence of the presentation.
- 01Scenario and simulation coverage
- 02Timing and control envelope
- 03Hardware-in-the-loop results
- 04Offline and degraded-mode behaviour
Current boundary
The limit is part of the result.
Diyar is a platform development programme. Hardware, control, safety, and environmental claims remain gated by the specific device and validation context.