Real-Time Fleet Intelligence Platform for a 3,200-Vehicle Logistics Operator
Challenge
SwiftHaul operated 3,200 vehicles across the UK with no real-time visibility. Dispatch relied on phone calls, ETAs were estimates, fuel usage was uncontrolled, and driver behaviour was unmonitored. Delivery ETA accuracy stood at 62%, impacting customer trust and operational efficiency.
Solution
We built a real-time fleet intelligence platform combining IoT telematics, a high-throughput streaming architecture, AI-driven route optimisation, driver behaviour analytics, and a customer-facing live tracking portal.
Results
ETA accuracy improved from 62% to 94%. Fuel consumption reduced by 11%. Driver incident rates dropped 43%. Customer satisfaction increased from 3.2 to 4.6/5.
SwiftHaul was operating a national logistics network without real-time visibility.
Dispatchers relied on phone calls to locate drivers. Customer ETA requests required manual follow-ups. Fuel spend was tracked after the fact, not optimised in real time. Driver performance varied widely with no consistent measurement framework. At 3,200 vehicles and growing at 15% annually, this model was no longer sustainable.
Deploying the Data Layer
The transformation began at the edge. We deployed 4G telematics units across all 3,200 vehicles over an 8-week rollout. Each device streams:
- GPS location every 10 seconds
- vehicle telemetry from the CAN bus
- behavioural signals such as harsh braking, acceleration, and idling
Installation was coordinated across regional depots, with parallel deployment teams ensuring minimal operational disruption. This created a continuous, high-frequency data stream across the entire fleet.
Real-Time Platform Architecture
At scale, location tracking is not a dashboard problem. It is a streaming systems problem. The platform processes 2.8 million location events per day through a low-latency pipeline:
- ingestion via MQTT brokers
- event streaming through Apache Kafka
- real-time processing using Apache Flink
- per-vehicle state updates computed in under 200ms
This architecture ensures that every vehicle’s position, status, and route is continuously updated and immediately available to downstream systems.
Route Optimisation Engine
Static routing was replaced with dynamic, data-driven optimisation. The routing engine combines:
- base routing via OSRM
- vehicle constraints (weight, height, restrictions)
- historical traffic patterns
- real-time traffic data
Routes are continuously recalculated as conditions change, improving efficiency without requiring dispatcher intervention.
Driver Behaviour Intelligence
Driver performance was previously invisible. The platform introduced a structured scoring system based on:
- harsh braking and acceleration
- speeding patterns
- idle time
- seat belt compliance
Each driver receives a daily score, feeding into a coaching programme managed at depot level. This turned behavioural data into actionable operational insight.
Customer-Facing Visibility
The platform extended beyond internal operations. Customers now receive:
- live tracking links
- real-time ETAs
- automated notifications via SMS and email
Delivery visibility became a product feature, not a support function. ETA accuracy is now contractually defined, strengthening client relationships and commercial positioning.
Measured Impact
The shift to real-time intelligence delivered measurable outcomes across operations:
- ETA accuracy: 62% → 94%
- Fuel consumption: reduced by 11%
- Driver incidents: reduced by 43%
- Customer satisfaction: 3.2 → 4.6 / 5
The platform paid back its implementation cost within 11 months through fuel savings alone.
Why It Worked
The impact came from three structural changes:
- Continuous data capture across the entire fleet
- Real-time processing, not batch reporting
- Actionable outputs embedded into daily workflows
This was not a reporting upgrade. It was an operational system redesign.
Final Thought
Logistics performance is determined by visibility. Without it, decisions are reactive. With it, operations become predictable, optimised, and scalable. The difference is not incremental. It is structural.
Building Fleet Intelligence Platforms?
Intagleo Systems helps logistics and transportation companies design real-time tracking platforms, optimise operations, and turn fleet data into measurable business outcomes.