AIMED Research Initiative
FUSE-EDGE
A self-contained simulation environment for exploring autonomous-system behaviour and edge cases in a clear, repeatable and easy-to-understand way.
What FUSE-EDGE Is Investigating
FUSE-EDGE explores how autonomous systems behave when they encounter unusual, ambiguous or failure conditions. The aim is to create controlled, repeatable scenarios where system behaviour can be observed, replayed and understood — making failures easier to investigate rather than simply recording that something went wrong.
Current research includes sensor behaviour, sensor disagreement and edge-case scenarios involving conditions such as poor visibility, occlusion and misleading visual information.
Research Pillars
Sensor Behaviour
Investigating how individual sensors respond under edge-case conditions — poor visibility, occlusion, misleading visual information — and what that means for system-level decision making.
Sensor Disagreement
Exploring scenarios where sensors produce conflicting outputs, and building tools to visualise and replay those disagreements in a way that is clear and interpretable.
Repeatability & Diagnostics
Creating controlled, repeatable failure scenarios that can be replayed and examined — making it easier to understand why a system behaved as it did, not just that something went wrong.
Test. Observe. Replay. Understand.
FUSE-EDGE is active research. Updates will be published here as the work progresses. If you're working on related problems and want to discuss the research, get in touch.
Contact AIMED