We don't just monitor — we predict.
BCM's Intelligent FMS is an AI-based operating system that goes beyond location monitoring to predict riders' credit and accident risk in advance. It is the first solution to port the automotive car-sharing standard to motorcycles.
The core solution that turns motorcycle driving data into enterprise operating decisions.
Five solution components
H/W (device) + S/W (monitoring, remote control, key-sharing) + AI (risk scoring) — the first integrated solution to port the automotive FMS standard to two wheels.
- 01 Core / Improved
FMS Device (H/W)
A rugged device that runs reliably in harsh two-wheeler environments (rain, vibration, heat). IP65+ dust/water resistance, targeting PASS on automotive standards 4.2.3 (vibration) and 4.2.6 (transient voltage), with ultra-low-power circuitry.
- 02 Core
Monitoring (S/W)
Per-second GPS tracking, ignition on/off status, and real-time battery monitoring.
- 03 Add-on
Remote control — restart prevention (S/W)
A safety design that keeps the current ignition on but blocks only the next start. Addresses non-payment/runaway scenarios while protecting rider safety — porting the automotive car-sharing standard to two wheels.
- 04 Add-on
Key-sharing (S/W)
Smartphone-based access control and usage authentication without a physical key. Optimized for multiple riders sharing one motorcycle.
- 05 ★ Core tech
Risk Management Engine & Scoring (AI)
Quantifies credit (non-payment) and accident (insurance) risk from multimodal data (driving, activity, contract, biometric). A hybrid of rule-based (regression) and learning-based (Autoencoder, Isolation Forest, XGBoost). TIPS target AUC 0.65+.
How it differs from automotive FMS
| Dimension | Automotive FMS (SOCAR, Lotte Rental, etc.) | BCM two-wheel FMS |
|---|---|---|
| Data | Mostly single driving stream | Driving + activity + contract + biometric (multimodal) |
| User | Transient users (hard to score) | Same user, high-frequency repeat (highly scorable) |
| Usage frequency | Occasional (a few times/month) | Constant (30,000+ km/year) |
| User identification | Weak | Strong — 1:1 rider matching |
| Scoring usefulness | Difficult (SOCAR's driver-score model failed) | Industrially meaningful |
A model validated on real data
Contract, activity, driving and accident-candidate data fused for B2B decisions.
Credit risk — 85-rider analysis
| Group | Count | Delinquency |
|---|---|---|
| Low (low risk) | 13 | 0.0% |
| Mid | 41 | 4.9% |
| High (high risk) | 31 | 16.1% |
In the High group, 4 of 5 showed concentrated, repeated delinquency compared with the Low group's cases (38.5%).
Accident risk — analysis of 151 real accidents
74%
Candidates detectable in advance from data
Precursor patterns derived from 31 intersection/turn/U-turn cases, 45 rear-end cases and 17 pedestrian/bicycle cases, among others.
Learn more
-
FMS Device (H/W)
IP65+ dust/water resistance, targeting PASS on automotive vibration & transient-voltage standards.
Details → -
AI Risk Scoring
Rule + Learning hybrid. Target AUC 0.65+ for credit/accident risk.
Details → -
Autonomous · e-Call
A 5-stage L1–L5 track. Automatic accident reporting and 119 integration.
Details → -
Enterprise SaaS
A B2B subscription for lease/rental firms, delivery agencies and insurers.
Details →