Fleet Tracking Dashboard

Dispatchers needed five tools and their phones to answer one question: where is this consignment? The company runs trucks across South India for manufacturers and distributors, and its operation had outgrown the way it was managed.

Trucks carried GPS devices from different suppliers, each with its own website. Trips and consignments lived in spreadsheets. Drivers reported progress by phone, proof of delivery was a paper slip that came back days later, and delays were usually discovered when the customer rang to ask. From April 2018, every consignment also needed a valid e-way bill, and an expired one on the road meant a stopped truck and a penalty.

We spent the first month in the dispatch room and at two hubs, sitting beside dispatchers through full shifts.

• Dispatchers spent their day on the phone, calling drivers for positions they could have seen on a screen.
• Every GPS supplier had its own website, so no one ever saw the whole fleet at once.
• Delays were only noticed when they were already late. Nobody knew what “normal” looked like for each route.
• Paperwork lagged behind the trucks. Lorry receipts, proof of delivery and e-way bills were chased separately from the trips they belonged to.

The dashboard brings every truck, trip and consignment onto one live map. The dashboard reads the GPS devices already fitted in the trucks, whatever their supplier, and shows each truck’s position, speed and status in real time.

Each trip carries its consignments and their lorry receipt numbers, with geofences around hubs, customer sites and common halt points. The dashboard knows when a truck reaches a hub, leaves a customer’s gate or stops somewhere it shouldn’t. Dispatchers filter the map by route, customer, branch or status, and see each truck’s whole day in one place.

A delay alert is only useful if it knows what normal is. The dashboard learns how long each route usually takes from past trips, including the usual night halts and loading times.

Every live trip is compared against that baseline, and the dashboard raises alerts in order of urgency:

• behind schedule: the truck is running slower than the route’s normal pace;
• unusual halt: the truck has stopped longer than usual, or somewhere unusual;
• at risk: the arrival time now falls after the time promised to the customer.

Alerts go to the dispatcher responsible for that route, so the team can act, by calling the driver, informing the customer or planning a handover, before anyone has to ask.

Drivers got a simple Android app with large buttons, in Tamil and English. It shows each trip’s consignments, addresses and contacts, and asks for only what matters: start trip, reached, unloaded, delivered.

At delivery, the driver takes a photo of the goods and records the receiver’s signature or a one-time code sent to the receiver’s phone. The consignment closes on the dashboard instantly. Highways often have no signal, so the app saves everything on the phone and sends it as soon as the network returns.

The national e-way bill system arrived in April 2018, in the middle of the project, and we built it in. Every consignment records its e-way bill number and validity, which depends on the distance travelled.

The dashboard warns the team before any e-way bill expires on the road, so it can be extended in time. Lorry receipts, e-way bills and proof of delivery are attached to the consignment they belong to, and month-end billing starts from complete delivery records instead of paper slips.

Every consignment has a tracking link that the company shares with its customer by SMS or email. The customer sees where the truck is, the expected arrival time and, once delivered, the proof of delivery, on any phone and without an account. Customers check for themselves instead of calling the dispatch room.

The dashboard is React, on a Node.js and Express backend with MongoDB for trips and positions and Redis for live state. A device gateway reads the different GPS suppliers’ data formats, and Socket.IO pushes positions and alerts to every dispatcher’s screen as they happen. Route timings are learned by Python models trained on past trips. The driver app is built in Kotlin for Android, Google Maps provides the maps, and everything runs on AWS.

Two dispatchers used the first live view alongside their old tools for a month, and their feedback shaped the alerts. The whole dispatch team then moved over, drivers were trained at the hubs as they came in, and the old spreadsheets were retired. In the years after, we added new routes and customers, and brought in FASTag toll records once FASTag became mandatory.

The dispatch team now runs the whole fleet from one screen, sees delays early enough to act, and closes deliveries the moment they happen. No e-way bill expires unnoticed on the road, customers follow their own consignments, and month-end billing starts from complete records.