A guided product tour
You cannot lose product you can weigh.
Follow one load from the plant gate to the customer's yard. Loadline puts a scale at every checkpoint, so the difference between what left and what arrived stops being an argument and becomes a number, attributed to a truck, a driver, and a trip.
Shrinkage hides in the gap between the plant and the customer.
A spreadsheet balances because someone made it balance. Bags go missing between the loading bay and the delivery point, and by the time the numbers are keyed in at month end, the trail is cold. The loss is real money, and nobody can point to where it went.
- Weights are typed by hand, so the record already agrees with itself before anyone checks the truck.
- There is no reading at the customer, so a short delivery looks identical to a full one.
- By reconciliation, the driver, route, and load are three weeks in the past.
| Document | Planned | Received | Gap |
|---|---|---|---|
| GRN 08812 | 12,000.0 | 12,000.0 | 0.0 |
| GRN 08813 | 9,500.0 | 9,500.0 | 0.0 |
| GRN 08814 | 17,000.0 | 16,700.0 | 300.0 |
| GRN 08815 | 6,250.0 | 6,250.0 | 0.0 |
kg unaccounted on GRN 08814. Which truck? Which stop? Which driver? The sheet cannot say. This is the question the rest of the tour answers.
A trip starts as a plan, and the plan is what the gate will check against.
Dispatch picks the site, the vehicle, and the driver, then adds the product lines. That plan sets the expected weight for the whole trip. Nothing about it can be edited once the truck is on the bridge, so the number the gate weighs is measured against a target set before anyone touched the scale.
Every trip crosses the scale twice, and the machine does the writing.
This is where Loadline earns its keep. Weighbridge indicators are read over their own APIs, so tare and gross land in the record untouched by a keyboard. Net is arithmetic, the verdict is automatic, and a load that misses tolerance stops at the gate until a supervisor signs an override with a reason.
- 1Gate-inThe truck is booked onto the bridge against its released trip.
- 2Weigh-in (tare)Empty weight read straight off the indicator, never typed.
- 3LoadBags counted on per customer against the load plan.
- 4Weigh-out (gross)Loaded weight captured. Net is gross minus tare.
- 5VerdictNet against plan and tolerance. Out-of-bounds loads are held.
- 6Gate-outThe moment, and the weight, it left the yard.
The gate-out weight, 17,000.0 kg, is now the load's truth on the record. Everything downstream is measured against it.
Between the gates, the trip keeps writing itself.
The driver carries the manifest on the phone and the route records itself as GPS breadcrumbs, with no signal required to keep working. The trail is the trip's alibi: where it went, how long it held, and whether it ever left the corridor between the plant and the delivery point.
Breadcrumbs queue on the device through dead zones and sync when the signal returns. A gap in coverage never becomes a gap in the record.
Illustrative route. In the console this is a live MapLibre fleet map; here it is drawn, not fetched.
The delivery is confirmed by a scale, not a signature.
Inside the customer's geofence, the received weight is captured and the customer confirms with a one-time code sent to their phone. Now there are two honest readings for one load: what left the yard, and what was received. Their difference is the shrinkage, and it lands the instant the code is entered.
- Weighed at the customer. Received weight captured on delivery, not estimated from a paper note.
- OTP confirmed. A one-time code proves the right customer accepted the right load.
- Geofence verified. Confirmation only counts inside the 200 m fence around the stop.
Above the ±85 kg tolerance, so the trip is flagged for investigation with the driver, vehicle, and route already attached.
One operator surface, every department's view of the same trips.
Dispatch works the board. The weighbridge team watches the live feed. Finance reconciles payments and files the tax register, and stores raise requisitions. Every screen reads from the same append-only trip record, so nobody is arguing from a different spreadsheet.
TR-1043 carries the 300.0 kg gap from chapter five, open with its evidence one click away.
The report your CFO has been asking the ERP for.
At the end of the month, shrinkage is not a mystery line. It is a table: kilograms out against kilograms confirmed, attributed by driver, vehicle, and route, ranked by who is losing the most. The conversation moves from whether product is disappearing to which corridor to fix first.
Every trip has two weighings and a confirmed receipt. Nothing is estimated, so nothing is unexplained. The losses that remain have a name.
| Driver | Route | Trips | Kg out | Kg confirmed | Shrinkage |
|---|---|---|---|---|---|
| Otieno | Nakuru | 18 | 214,000.0 | 213,900.0 | 0.05% |
| Wanjiru | Thika | 22 | 261,500.0 | 261,180.0 | 0.12% |
| Mwangi | Machakos | 16 | 192,000.0 | 190,900.0 | 0.57% |
| Kamau | Kitengela | 20 | 238,000.0 | 237,760.0 | 0.10% |
Machakos runs an order of magnitude hotter than the fleet. That is where the 300.0 kg came from, and where the next audit goes.
Put a scale at every checkpoint.
Loadline runs on your weighbridges, your vehicles, and your routes. Start with one plant and one product line, and watch the first unexplained trip get a name.