ADAS Health Tier Semantics for Fleet Operations
NADIR maps continuous sensor residuals to NOMINAL, CAUTION, and CRITICAL tiers so fleet safety officers, MSO partners, and insurers share one vocabulary. This guide explains tier transitions, shadow advisory guardrails, Console workflows, and signed evidence exports — the operational contract behind NADIR calibration intelligence.
Why tiers beat raw scores in fleet ops
Engineers love continuous metrics; executives need decision thresholds. Tiers translate multivariate residuals into actions: monitor, schedule inspection, or halt dispatch for a VIN. Shadow pilots validate alert rates before operational holds — see shadow-mode pilots.
Mahalanobis and conformal methods inform band edges; operators read Console histograms. Deep dives: Mahalanobis fleet health and conformal prediction monitoring.
Console, API, and evidence alignment
Console shows tier history per VIN; API routes expose engineer-grade residuals for BI. Signed bundles attach tier timelines to repair and insurance workflows — documented on evidence exports and in MSO post-repair evidence.
Maintenance scheduling hooks connect tiers to CMMS tickets — see fleet maintenance calibration scheduling.
Regulatory and insurer context
FMVSS 127 and 2026 liability trends increase demand for provable sensor stewardship. Tier history reduces discovery cost when ADAS performance is disputed — cross-link FMVSS 127 compliance and fleet insurance liability.
Escalation playbooks safety boards expect
Define CAUTION rate bands before pilots begin — not after the first alert wave. Document review cadence, weather exclusions, and shop escalation paths in LOI exhibits. Week-two review boards separate environmental noise from extrinsic shift using modality breakdowns available in Console.
Fleet legal teams appreciate tier vocabulary because it avoids deterministic claims OEM stacks cannot support. Graduation to operational holds requires explicit stakeholder sign-off captured in evidence metadata — never implied by dashboard color alone.
Pair with fleet ADAS compliance and Detroit ecosystem context in Detroit ADAS infrastructure.
Cohort learning across duty cycles
Highway-heavy duty cycles produce different baseline variance than urban last-mile — cohort tags prevent false CAUTION on route assignment alone. Maintenance windows should align with tier spikes: glass campaigns, seasonal soiling, post-collision clusters.
Insurers prefer structured tier history over ad hoc screenshots — reduced discovery cost when ADAS performance is disputed pre-incident. MSO partners attach the same bundles to franchise scorecards.
Micro-pilots under fifty vehicles may cohort by platform until per-VIN depth accumulates — see micro-pilot ROI.
API and Console contract surfaces
Operators live in Console; data science lives in API exports. NADIR keeps both aligned on tier timestamps so BI dashboards never disagree with dispatch holds. Engineer routes expose modality contributions; executive routes stay histogram-first.
Webhook and batch export patterns mirror telematics ingest — same org isolation, same idempotency keys. Read API reference before wiring production keys.
FAQ
Can tiers differ by OEM platform?
Yes — org-scoped models normalize ingest profiles for mixed fleets.
When do tiers trigger dispatch holds?
Only after explicit stakeholder sign-off; shadow mode scores without dispatch changes by default.
Are tier definitions public?
NADIR publishes semantics sufficient for legal review without exposing proprietary coefficients.
Who wrote this guide?
Dhruv Hegde, Co-Founder and CEO at NADIR.
Next steps
Review the NADIR platform, explore the HORIZON pilot walkthrough, and open the Calibration Lab before wiring fleet telemetry. Shadow pilots score every eligible VIN without changing dispatch.
Request a four-week cohort via the footer pilot form or team@nadirai.net with fleet size, telematics partner, and target KPIs.