GPS stitched to the ERP: the ETA that updates itself

The problem

Two modern systems, connected by a dispatcher with two windows open.

It's a relatively modern system, with an available API, but "integrating it properly takes time" — so nobody has done it yet. The dispatcher opens the GPS platform, finds the truck, eyeballs the ETA, and re-types it by hand into the management system or an email.

• The operator can see where each of its dozens of tractor units and trailers is at any moment on the fleet GPS platform. The data exists, live, for every truck.
• The platform has an API, but “integrating it properly takes time”, so nobody has done it yet. It is always next quarter's project.
• The dispatcher opens the GPS, finds the truck, eyeballs the ETA and retypes it by hand into the management system or an email, several times a day and for every delayed delivery.
• Meanwhile clients call or write to ask where their delivery is, and customer service asks the dispatcher again, who opens the GPS one more time.

How it fits the IRIS system

Connect stitches the GPS and the ERP through their APIs — no double entry.

Connect links the two modern systems via API: the position and ETA of every shipment flow automatically into the matching order. If the customer has a portal, the updated ETA arrives there directly, with nobody at the operator having to type it in.

This case is not about reading paper or screens; both systems are modern. It is about the gap between them, which today is filled by a person copying an ETA. That gap is exactly what Connect closes, and the dispatcher goes back to the work only a dispatcher can do: handling the exceptions, the breakdowns and the slots that are about to be missed. With 30 tractor units on the road, the minutes saved per ETA add up to a large part of a dispatcher's day.

See the full IRIS architecture →

Before and after

Today's ETA versus the stitched ETA

Aspect	Today	With iLEAN Connect
How the ETA reaches the order	Retyped by the dispatcher	Flows from the GPS automatically
Updates per shipment	When someone has time	Continuous
Customer portal	Shows the last typed ETA	Shows the live ETA
Where-is-my-truck calls	Answered by hand	Fewer, the portal answers
Dispatcher's time	Copying between two windows	Spent on exceptions
The GPS platform	—	Kept as it is

Before: the dispatcher checks the GPS and re-types the ETA by hand into another system, several times a day. After: the ETA flows automatically from the GPS to the order and the customer portal.

Impact estimate

Impact estimate — to be validated with your numbers.

The block below is an estimate to be validated against your plant's actual data. We put it forward so the committee has an order of magnitude; we refine it during the assessment.

• Estimated payback 3-6 months, the shortest of the matrix, because both systems already exist and only the stitch is missing.
• An always-updated ETA on every order, with no double entry by the dispatcher.
• Fewer tracking calls and emails to customer service, which today interrupt both customer service and dispatch.
• And better perceived service with no added headcount, because the dispatcher stops being the integration between two systems.

Payback 3-6 months · always-updated ETA with no double entry. Fewer tracking calls and emails to customer service, and better perceived service with no added headcount.

And the fair question from the production manager

“What if the ETA it posts is wrong and the client plans around it?” — the ETA is not invented: it comes from your own GPS platform, and matching each vehicle to its order is an anchored task, where the best models drop below 1.5% error [1]. When a truck cannot be matched unambiguously to an order, it is shown to the dispatcher instead of being posted. And the dispatcher can still override any ETA manually; the system records that it was overridden and by whom.

^[1] OpenAI paper "Why Language Models Hallucinate", 2025 — on the reliability of AI in anchored tasks.