Every 5 minutes for clocked-in workers, analyses the last 60 minutes across six risk axes to score risk and level, and alerts by priority before it turns into an accident.
SenseZero AI Risk Prediction analyses clocked-in workers every 5 minutes: the last 60 minutes of data from BLE beacons and connected wearables is scored across six risk axes into a 0–100 risk score, and at 60 or above managers receive the P0–P4 priority with level-specific immediate actions.
How it works
A 3-stage AI pipeline from data collection through to corrective action.
Data from site devices and the app is collected continuously in 60-minute windows.
A weighted judgement across six risk axes; axes without devices are excluded.
Each level comes with on-site actions to take right away.
Key features
Analyses clocked-in workers every 5 minutes, using hazard-zone entries, PPE wearing, beacon detection and (with a connected watch) biometrics to produce a 0–100 risk score.
Finds workers who repeatedly enter hazard zones or whose beacons go undetected for long, and reflects it in their risk score and priority. The control center shows each zone's workers and assigned managers.
Five priority levels are set automatically and sent to managers together with the on-site actions to take for each level.
Risk engine
Risk is judged by combining the weights of six risk axes. Weights are relative; axes for devices that are not deployed are excluded and the rest renormalised. Data completeness and quality are not added to the score but shown separately as data confidence.
| Axis | Relative weight |
|---|---|
| Data completeness (confidence) | Not scored |
| Data quality (confidence) | Not scored |
| PPE wearing events | 20% |
| Hazard-zone entry | 15% |
| Beacon non-detection count | 10% |
| Beacon non-detection duration | 10% |
| Biometric composite | 10% |
| Trend analysis | 10% |
Priority
Sets one of five priority levels automatically and sends the on-site actions to take for each level.
| Level | Category | Immediate actions | Triggers |
|---|---|---|---|
| P0 | Emergency response | Stop work · dispatch · call 119 if needed | SOS · fall detection |
| P1 | Biometric risk | Stop work · rest and re-measure | Abnormal heart rate · stress · SpO2 |
| P2 | PPE & zone risk | Check PPE · recurrence-prevention steps | Accumulated PPE non-wearing + high-risk zone |
| P3 | Beacon non-detection | Check beacon wearing · battery · connection | Accumulated / prolonged non-detection |
| P4 | Repeated zone exposure | Minimise high-risk zone stay · spread work | Repeated high-risk zone exposure |
While a worker is clocked in with the app running, a risk score is calculated every 5 minutes. Axes for devices you don't have are excluded, so it works with any device mix.
SenseZero wearable and beacon data is the primary input; integration with external systems can be customised via API.
Each result and its detected indicators are kept in the AI risk analysis history for follow-up review.
Accident prevention through early discovery of risk patterns
Every priority comes with actions to take right away
Safety decisions based on data, not experience alone
Hazard-zone entry history and beacon detection from BLE beacons, helmet and harness wearing events, and biometric signals such as heart rate and SpO2 from connected wearables are collected over a 60-minute window (biometrics: last 10 minutes) and scored across six risk axes every 5 minutes while on shift. Axes for devices you don't have are excluded.
Five levels are set automatically: P0 (SOS or fall), P1 (abnormal biometrics), P2 (accumulated PPE non-wearing in a high-risk zone), P3 (accumulated or prolonged beacon non-detection) and P4 (repeated high-risk zone exposure). Managers receive the matching immediate actions with each alert.
The current AI risk prediction engine primarily uses SenseZero wearable and beacon data. Custom integration with external systems is available via API; specific requirements can be scoped after a demo session.
While clocked in, risk scores are calculated every 5 minutes, and results appear in the dashboard's AI risk analysis history. The trend axis looks at recent biometric changes, so its evidence grows as data accumulates.
We will propose the optimal solution combination tailored to your industry, site scale, and management goals.
