A live-detour loop that treats the recommended route as a treatment assignment, and infers blockage only when an assigned approaching cohort exists.
Question
Speed-based maps treat silence as flow. If nobody is on the recommended route, an empty road and a blocked road look the same. The empty-road paradox is not a data gap. It is an identification failure: the system is not entitled to the conclusion it wants to draw.
Constraint
A navigation system may claim blockage only when an assigned approaching cohort exists. No cohort, no treatment, no inference. Silence is not evidence of a clear road, and it is not evidence of a closed one. The admissible action is to wait.
Method
The recommended route is a treatment assignment, not a sensor. Approaching-cohort queries ask whether anyone assigned to that treatment is actually on it. When a cohort exists and delay appears, a live detour braid is first-class: not a dashed alternative, but a closed-loop reroute with its own identification conditions. Monte Carlo experiments compare the loop against silence and exploit baselines.
Artifact
A research prototype of ICCN — identification-constrained closed-loop navigation — with approaching-cohort queries and live detour braids.
Ongoing doctoral research, Department of Computer Science, Kean University.
Identification constraint
Not identified
Assigned route · waiting
No assigned approaching cohort. The recommended route is a treatment with no units. Speed-based maps would call this a clear road. ICCN refuses the conclusion. Silence is not flow.
Schematic, not a traffic model. The claim is about identification: when a navigation system is entitled to infer blockage, and when it must wait.