A frontier proposal with a validation plan, not an implemented detector.
Intended capability
Research whether the field could support this proposed outcome: Detects the fleet's physical hallucinations as repetition pathologies: loop hallucination — the same zone-transit string recurring with machine-exact period and no downstream progress; path thrash — sustained motion energy oscillating between adjacent zones with near-zero net displacement; and dwell waste — flow collapse and queue growth in a lane the estate declares actively tasked. Each candidate is a physics statistic with a lawful twin: patrols are cyclic, obstacle avoidance oscillates, buffers idle by design. The reasoning tier consumes the candidate plus the declared task class and renders the waste-or-legitimate judgment, naming the twin it rejected in every finding.
This describes the intended outcome. Readiness is research concept, evidence is class C, and a catalog mapping or recording is not proof of this outcome at a real site.
Solution blueprint
See the environment before installing it.
This exact kit is one of 191 first-class designs. It includes geometry, objects, nodes, wording, scenarios, installation, limitations, and catalog-bound readiness.
No rendered revision is available yet.
Bundled recording
A related Fleet scene
This is one recorded vertical scenario. It is not separate validation of every capability in the catalog.
01 The physics
CSI owns minute-scale trajectory statistics: transit strings across the zone graph, loop-period variance (machine loops repeat with a regularity no human route sustains), motion-to-displacement ratio between adjacent node cells for thrash, and the flow-null contradiction for dwell — a still robot is as RF-dark as a still human, so dwell is observed as the upstream queue plus the declared-tasking mismatch, never the idle machine itself. Fine approach-retreat geometry inside a single zone belongs to an aimed mmWave cone, not the field.
02 Shared processing path
Zone-transit string extraction -> recurrence scoring + loop-period variance (loop) -> reversal rate + motion-to-displacement ratio between adjacent cells (thrash) -> per-lane flow statistics + queue-growth detection against declared tasking (dwell) -> reasoning-tier classification against declared task class and zone semantics, twin named per verdict -> waste findings, class-C tagged, into the certificate stream.
This is a capability design path. Components may be shared with other catalog entries; it is not presented as a unique algorithm.
03 Validation plan
Own-hardware protocol: staged failure library on a test fleet — scripted loops, forced replanning oscillation, planted lane blockages — interleaved with legitimate patrols and buffer idles; score candidate detection and reasoning-tier twin separation against the script. Class C until detection rates are calibrated.
04 Commercial hypothesis
'Hallucination Ledger' — for the fleet operations owner who suspects the capacity gap is robots arguing with their own maps: an independent count of the cycles your fleet burned going nowhere, judged against the work it declared.