How to Add Unloading Detection to Any GPS Tracking System

GPS tracking is table stakes in bulk material logistics. If your fleet runs cement, coal, aggregates, or any bulk commodity, there's a very high chance you already have GPS trackers on every truck. The data tells you where each vehicle went and when it arrived at each location.

What it doesn't tell you is what happened after arrival. Did the truck unload? When did it start? When did it finish? Was it a full discharge or partial? Did the unloading happen at the right location?

This is the gap that activity sensing fills — and the key point is that it fills it as an add-on layer, not a replacement. You keep your GPS system. You keep your telematics hardware. You add activity sensing on top, connected via the data feed your existing system already emits.

Why Standard GPS Can't Detect Unloading

GPS tracks position. It can tell you:

The truck stopped at coordinates X,Y at time T

The truck was stationary for 47 minutes

The truck then departed

It cannot tell you:

Whether the truck discharged material during those 47 minutes

Whether the discharge was complete or partial

Whether the discharge happened at the correct facility (not just the correct geofence area)

Geofencing gets you one step closer — you know the truck was inside a defined delivery zone. But a truck can enter a geofence, remain stationary for 47 minutes without discharging anything, and exit. The geofence trigger fires either way.

Activity sensing uses the vehicle's behavioral signals — combined with the AI classification model — to answer what the truck was actually doing. This requires sensor data beyond GPS position alone, but critically, most modern telematics devices already collect this data. It's just not being used.

What Data the Activity Sensing Engine Uses

Modern telematics hardware collects far more than GPS position. A typical Samsara, Geotab, or Verizon Connect device emits:

GPS position (lat/lng, speed, heading)

Accelerometer data (motion, vibration, impact)

Engine telemetry (ignition state, RPM, PTO engagement)

Battery voltage

Temperature (if equipped)

The activity sensing engine ingests this full telemetry stream — not just GPS — and applies a trained classification model to determine what the vehicle is doing. The accelerometer and PTO data are particularly valuable: the vibration signature of a running pneumatic cement tanker compressor is distinct from a stationary idle. The PTO engagement pattern of a ready-mix drum running during discharge is distinct from transit.

The Integration Architecture

Step 1: Connect Your Telematics Data Feed

Every major telematics provider offers a data export or API. The activity sensing engine connects to your existing provider's data stream:

For older GPS units that don't offer a real-time API, position data can be fed via scheduled FTP upload or email attachment. Activity sensing accuracy is somewhat lower for batch-fed data vs. real-time streams, but core unloading detection remains functional.

Step 2: Vehicle Type Calibration

Activity signatures differ by vehicle type. A rear-discharge concrete mixer looks different from a pneumatic cement tanker. A side-tipper aggregate truck looks different from a walking-floor trailer. During the calibration phase (typically 5–10 business days), the classification model is trained on your specific vehicle types:

A sample of confirmed loading events is provided (trips where loading definitely occurred)

A sample of confirmed unloading events is provided

The model identifies the distinguishing signal patterns for your fleet

A validation run is performed on a held-out data sample to confirm accuracy

Step 3: Delivery Manifest Integration

To enable authorized vs. unauthorized discharge detection, the activity sensing engine needs to know where deliveries are supposed to happen. This is connected via:

TMS dispatch data (automated API feed of planned delivery locations)

Manual delivery manifest upload (CSV)

Geofence library (list of authorized facility locations and radii)

Step 4: Alert and Event Configuration

Define the rules for what generates an alert vs. what's logged silently:

Unauthorized discharge → immediate alert to dispatch manager

Partial discharge detected → logged, optional alert if below threshold %

Extended idle (loaded) outside facility → alert after configurable minutes

Unloading confirmed at destination → logged, TMS record updated

Step 5: Output Integration

Activity events are delivered to wherever your team needs them:

API: Real-time event stream for TMS or custom systems

Webhook: Push notification to your endpoint

Dashboard: Activity sensing portal for dispatch and operations teams

Reports: Daily/weekly delivery verification reports

Email/SMS: Alerts for critical events (unauthorized discharge, extended idle anomaly)

What Changes — And What Doesn't

Drivers don't need new devices, new apps, or new workflows. Dispatch teams get more data in the systems they already use. IT teams don't need to manage new hardware or telemetry infrastructure.

Industries Where This Integration Is Most Impactful

Cement & Ready-Mix — Grey market prevention, delivery verification, partial discharge tracking for multi-drop routes.

Coal — Unauthorized discharge detection on mine-to-plant routes, delivery verification at power plant yards.

Aggregates & Crushed Stone — Per-drop job site verification, short-loading detection, driver productivity benchmarking.

Fly Ash & Industrial Byproducts — Quality-sensitive delivery verification, batch traceability.

Fertilizers & Agricultural Minerals — Seasonal delivery verification across distributed dealer networks.

Getting Started

If your fleet already has GPS or telematics hardware from any major provider, you can have activity sensing data flowing within 5–10 business days. No procurement process for new hardware. No truck downtime for installation.

Talk to our team about connecting your existing GPS system to Intugine's activity sensing engine.