AI-Powered Predictive Maintenance for Vehicles
Predictive maintenance in the automotive industry covers two areas: fleet management and automotive service networks (dealers, service stations). ML approaches reduce unplanned downtime by 25-40% and optimize maintenance costs by transitioning from interval-based to condition-based servicing.
Data Sources
CAN-bus and OBD-II telemetry:
can_data_channels = {
'engine_rpm': 'OBD PID 0x0C',
'vehicle_speed': 'OBD PID 0x0D',
'coolant_temp': 'OBD PID 0x05',
'engine_load': 'OBD PID 0x04',
'fuel_trim_short': 'OBD PID 0x06',
'fuel_trim_long': 'OBD PID 0x07',
'intake_manifold_pressure': 'OBD PID 0x0B',
'dtc_codes': 'OBD Mode 0x03', # diagnostic fault codes
'oil_temp': 'OEM extended PID',
'transmission_temp': 'OEM extended PID'
}
Telematics devices (GPS + CAN): Teltonika, CalAmp, Webfleet Solutions (TomTom) — devices for fleets. Frequency: 1-10 sec. Data: coordinates + CAN parameters → cloud platform.
Dealer data:
- Repair history by VIN (from DMS — Dealer Management System)
- Warranty claims: repeat repairs = sign of incomplete fix
- PDI (Pre-Delivery Inspection) data
Failure Prediction
Component-based approach:
Brake pads:
def brake_pad_remaining_life(brake_thickness_mm, driving_style_features,
road_conditions, mileage_km):
"""
Regression model: residual brake pad life
Features: thickness, driving aggression, urban percentage
"""
features = np.array([
brake_thickness_mm,
driving_style_features['hard_braking_events_per_100km'],
driving_style_features['avg_deceleration'],
road_conditions['urban_pct'],
mileage_km
])
remaining_km = brake_wear_model.predict([features])[0]
return remaining_km
Battery (12V and HV for EVs):
- SoH (State of Health) from starting voltage and load
- Internal resistance: grows with degradation
- Cold cranking amps (CCA): failure prediction at low temperatures
Engine — early signs:
- Long Term Fuel Trim > ±10% → rich/lean mixture
- RPM fluctuations at idle → spark plugs, ignition coils
- Compression loss → piston ring wear (requires compression test)
DTC analytics:
def dtc_risk_score(dtc_history, vehicle_profile):
"""
DTC codes as degradation indicators:
P0300-P0312: misfires → spark plugs/injectors
P0420: catalytic converter below threshold
U-codes: CAN-bus communication errors → wiring
"""
recurring_dtcs = find_recurring(dtc_history, min_occurrences=2)
risk_by_system = classify_by_system(recurring_dtcs)
return risk_by_system
Fleet Management
Fleet telematics:
Daily health score per vehicle:
def fleet_vehicle_health(vehicle_id, last_7days_telemetry):
features = aggregate_telemetry(last_7days_telemetry)
# Behavior anomalies
anomaly_score = isolation_forest.predict([features])
# Component wear
component_scores = {
'brakes': brake_model.predict(features),
'battery': battery_model.predict(features),
'engine': engine_model.predict(features)
}
overall_health = np.mean(list(component_scores.values()))
return {'health': overall_health, 'components': component_scores, 'anomaly': anomaly_score}
Maintenance optimization in fleet:
- Calendar schedule: minimize simultaneous downtime (not >15% of fleet)
- Just-in-time maintenance: when, not by mileage
- Spare parts: pre-ordering based on replacement forecast → reduce storage costs
Automotive Service Network (Dealer Use Case)
Proactive Service Campaigns: OEM-dealer + telematics → proactively invite customer for service before problem appears:
- "Your vehicle shows signs of brake pad wear. We recommend inspection at next service."
- ML-model trigger → CRM task → email/SMS to customer
RO Prediction (Repair Order): Forecast needed work before visit. Service advisor sees recommended list before car arrives:
ro_prediction_features = {
'vin': vehicle_id,
'mileage': current_odometer,
'last_service_items': last_ro_items,
'telemetry_flags': active_dtcs + wear_flags,
'months_since_last_visit': calendar_delta
}
predicted_ro = ro_model.predict_required_jobs(ro_prediction_features)
Parts Pre-Positioning: Based on aggregated RO forecasts for appointments next week — auto-order parts for dealer stockroom 2-3 days ahead.
Integration
DMS (Dealer Management System): CDK Global, Reynolds & Reynolds, 1С:Dealer — API for task and work order creation.
OEM platforms: Mercedes-Benz ME connect, BMW ConnectedDrive, ŠKODA Connect — branded telematics systems with open APIs for dealers.
Aftermarket: CARFAX, AutoVIN — enriching service history from independent sources.
Timeline: OBD-II connector + basic wear indicators + fleet dashboard — 4-5 weeks. ML component forecasting + DTC analytics + DMS integration + proactive campaigns — 3-4 months.