Predictive Maintenance in 90 Days: CAN/OBD Signals that Save Real Money (2025)
Read time: ~16 min
Last updated: 30 Oct 2025
From raw fault codes to fewer breakdowns. This whitepaper shows how to baseline signals, prioritize DTCs, and run a 30/60/90 plan that cuts cost per km.
Executive insight: Fleets commonly reduce unplanned breakdowns by 25–40% and maintenance cost/km by 8–12% in the first 90 days.
Executive summary
- Predictive maintenance works when data, policy, and people connect: signals → triage → action → KPIs → finance.
- Start with a small set of high-value signals (ECT, oil pressure/temp, battery voltage, misfire counts, DPF soot %, DEF).
- Use a priority map for DTCs so the shop knows what gets attention first.
Signals that matter (first 90 days)
Begin with 10–12 signals that predict real failures or fuel loss. Keep scope focused while you stabilize data quality.
| Signal | Why it matters | Quick rule |
|---|---|---|
| ECT (coolant temp) | Overheat & cooling health | No unexplained spikes; watch hot climates |
| Oil pressure & temp | Lubrication & bearing safety | No drops under load; match temp |
| Battery voltage | Starter/alternator health | > 10V during crank; 13.8–14.5V running |
| Misfire count | Ignition/fuel faults | < 3 per 1k cycles; investigate if rising |
| Fuel trims (STFT/LTFT) | Combustion efficiency | Stay within ±5% (goal ±3%) |
| DPF soot load % | Regen health & backpressure | Stable trend; avoid frequent forced regens |
| DEF level/quality | Emissions compliance | No quality warnings; match NOx/temps |
| TPMS (per wheel) | Rolling resistance & safety | ±2 PSI from spec; heat expansion awareness |
DTC priority map
Not all faults are equal. Prioritize DTCs to protect the engine, emissions systems, and safety.
| DTC | Description | Severity | Action | Clock |
|---|---|---|---|---|
| P0300 | Random misfire | High | Pull vehicle; inspect coils/plugs/injectors | 24h |
| P0118 | ECT high input | High | Stop & diagnose cooling system | Immediate |
| P2002 | DPF efficiency | High | Regenerate/inspect leaks | 48h |
| P0420 | Catalyst efficiency | Medium | Check O2 sensors & leaks | 7d |
Baseline & policy (first 2 weeks)
- Verify data quality on the pilot group (10–20 vehicles).
- Import 12 months of service history and parts spend.
- Publish the Predictive Maintenance Policy with roles and KPIs.
- Define the triage queue: who receives alerts, when, and how they close the loop.
30/60/90 plan
0–30 days — Stabilize data & triage
- Confirm streams (ECT, oil, voltage, DPF/DEF, misfire).
- Start a daily triage queue with severity rules (High/Medium/Low).
- Set up PM schedule per platform (Service A/B/C intervals).
31–60 days — Close the loop
- Measure repeat fault rate (fix quality).
- Add parts lead-time tracking and build a fast-mover list.
- Weekly reliability stand-up: PM%, MTTR, repeats, breakdowns.
61–90 days — Optimize & report
- Reduce false positives; refine thresholds by model/year.
- Publish a finance readout: breakdowns/100k km, cost/km, ROI.
- Roll out to remaining depots; lock the process.
KPI pack (what to track)
Profile: 120 vans, urban last-mile, high phone-use baseline.
| KPI | Definition | Target | Notes |
|---|---|---|---|
| Unplanned breakdowns / 100k km | Immobilizations normalized by distance | < 0.4 | Main reliability KPI |
| PM compliance % | Planned maintenance completed on time | ≥ 90% | Foundation metric |
| Repeat fault rate % | Same DTC within 14 days | < 10% | Measures fix quality |
| MTTR (hours) | Time from open to close | < 48 | Exclude parts wait |
| Parts lead-time (days) | Requisition to receipt | < 5 | Stock fast-movers |
| Cost per km (maintenance) | Spend normalized by distance | -8–12% in 90 days | Blended goal |
Case study — Mixed delivery fleet
Profile: 70 vehicles; city + regional; diverse platforms.
| Metric | Before (60 days) | After (next 60 days) | Delta |
|---|---|---|---|
| Breakdowns / 100k km | 0.72 | 0.41 | -43% |
| Repeat fault rate % | 22% | 9% | -13 pp |
| MTTR (hours) | 62 | 44 | -18 |
| Cost per km (maintenance) | 0.41 AED | 0.36 AED | -12% |
| Parts lead-time (days) | Requisition to receipt | < 5 | Stock fast-movers |
| Cost per km (maintenance) | Spend normalized by distance | -8–12% in 90 days | Blended goal |
- Levers: ECT/oil alarms, DPF priority handling, faster parts, PM compliance gating.
💬 FAQs
Do we need advanced sensors to start?
No; start with OBD-II/CAN data available today.
What about older vehicles?
Use add-on sensors or limit to available channels; focus policy & PM compliance.
How do we avoid alert fatigue?
Prioritize High; route Medium to PM; summarize Low weekly.
Will drivers resist shop visits?
Show repeat-fault and breakdown risk; schedule smart to avoid peak hours.
Author & Review
Author: V Zone International — Maintenance Analytics Team
Version: 1.0 • Last reviewed: 30 Oct 2025
Share
LinkedIn
X