Proposal for the Vos Logistics ETA project team

All the data already exists at Vos.
Procai makes it work — one tool at a time.

GPS, tachograph, Lovos planning, customer opening hours — it’s all there today. The first sprint turns it into one thing, done properly: an automatic ETA your CS agents, planners and fleet managers see at a glance. Delivered in twelve weeks. Owned by Vos, running in Vos systems. Everything else in the demo shows where that foundation can take you next.

12 weeks
from kick-off to a working, judged automatic ETA
100%
owned by Vos — code, docs and data stay in your systems
€0
subscription or licence fees after delivery — no lock-in
1
fixed senior person on your project, start to finish
Why this project, why now

Twenty years ago one fleet manager handled 40 drivers. Today it’s 20–25 — while more data is available than ever.

The ETA field in Lovos already exists. The tachograph already knows the remaining drive time. The board computer already knows where every truck is. Customer opening hours are already registered. What’s missing is not data, and not another big system — it’s the piece that connects what you have and puts the answer in front of the right person, automatically.

“A truck leaves Amsterdam today. What time is it in Milan tomorrow?” — that should never be a question a human has to calculate, type in, or chase by phone.

Our promise is deliberately narrow: one project, finished, in production, embraced by its users. That first visible success is what creates the momentum for everything after it — not a three-year masterplan.

Why Procai

The conditions you set for a partner — built into how we work

🔑

You own everything. No lock-in.

The code runs in Vos systems, not on our servers. No subscription, no licence, no dependency. Everything is documented so well that Vos IT — or any other party — can take over tomorrow. If we ever stop being the best choice, you lose nothing.

Small start, real proof

Sprint 1 is deliberately pure: only the automatic ETA, working and judged in 12 weeks, with a fixed, agreed scope and price. You see it work — with real Vos data — before any bigger commitment. No endless trajectory, no project that is “done in three years”.

🧭

Built with the end goal in mind

The ETA module is designed as a foundation, not a dead end: the same data stream later powers unload detection, waiting costs, document checks and customer mails. No boxed-in solution that blocks the next step. No big tangle.

💡

A partner that thinks along

We learn your processes and come back with “did you know this is also possible?” — not just execute a ticket. The proactive signals in the demo (waiting detection, opening-hours advisories) are exactly that kind of thinking, shown instead of promised.

🤝

Side by side with Vos IT

Clean split: Procai builds the external calculation module; presentation of fields inside Lovos stays with your IT team, connected via API. During development we only read data out of Lovos — nothing is written in until it’s proven and approved.

🗣

Honest about capacity

We are a small senior team — an AI specialist and an experienced developer (banking & warehouse-automation background), with trusted freelance developers when needed. That means fast switching and low overhead. It also means we will tell you ourselves if something doesn’t fit, before it becomes your problem.

The route — one building block at a time

Each step delivers value on its own. Each step makes the next one easier.

This is the opposite of a big-bang programme: every block below is a small, testable project with its own finish line. The ETA comes first because almost everything else depends on it.

To be clear: the first sprint — and the only thing we propose committing to now — is step 1, the pure automatic ETA. Steps 2–5 show where that foundation can take Vos in further development; each would be its own small project, started only after the previous one is finished and embraced.
Step 1 · the first sprint — this proposal

Automatic ETA in Lovos  → see it working

ETA + remaining drive time per trip, computed from GPS, tachograph and planning — visible to CS, planners and fleet managers without anyone typing it in. Honest ±15–30 min bandwidth. Finished, judged and in use within 12 weeks.

uses: Lovos · board computer · tachograph · routing API
Step 2

Proactive load/unload signalling

The system notices a truck standing at an unload address, sets the status, asks the driver “are you being unloaded?”, logs deviations and alerts CS — before anyone has to spot it.

uses: the ETA foundation + geofencing + driver messaging
Step 3

Waiting costs, captured and billable

Waiting time recorded automatically with evidence; CMRs scanned and read by AI (even crumpled ones — same technology as our receipt recognition for accountants). CS approves charge-on with one click, substantiation attached. No more 4-hour discussions about 2 hours of waiting.

uses: step 2 signals + CMR document AI
Step 4

Document checks before invoicing

Missing POD or CMR copy? Unusable scan? The system identifies the driver and chases the document automatically — invoicing without delay, without manual follow-up.

uses: step 3 document AI
Step 5

Automatic answers to “where is my truck?”

Customer mails answered automatically from the live ETA; delay notifications sent proactively; complex cases escalated to a human. The same pattern works for the creditor mailbox (“when will my invoice be paid?”).

uses: step 1 ETA + e-mail AI (proven at an accountancy firm)
Parallel quick win

Driver-language auto-translation

A small connector between Lovos/Trans and the board computer: planner writes in English, the driver reads it in Georgian, Uzbek or Russian — and vice versa. No waiting for the supplier; solvable in roughly a week.

independent of the ETA — can start any time
The 12-week first sprint

From kick-off to a working ETA your team can judge — nothing more, nothing less

Fixed scope, fixed price, one deliverable: the automatic ETA, live in Lovos. Users are involved in what we must achieve — not asked to design the solution. The other functions in the demo are explicitly not part of this sprint.

Weeks 1–2

Data & access

Read-only API access to Lovos, board-computer and tachograph data with Vos IT. Map the fields against the concept your team already wrote.

Weeks 3–5

First calculation

ETA engine v1 on historical trips: position + route + drive/rest rules. Measure accuracy against what actually happened.

Weeks 6–9

Live & honest

Running on live data, shadow-mode. Bandwidth tuned per route type. Edge cases from planners and fleet managers fed back in.

Weeks 10–11

Into Lovos

ETA delivered via API into the existing ETA field, together with Vos IT. Screen presentation by your IT team.

Week 12

Judge it

Side-by-side review with the project team: accuracy report, full documentation, and a go/no-go you can defend to the steering group.

Proven elsewhere

The same building blocks, already in production

2 FTE → 2 h/day

Compliance document analysis — asphalt industry

Research reports of 500–1,000 pages checked against legislation automatically, with source references in a dashboard. Two full-time specialists now spend two hours a day supervising instead.

same pattern as: CMR & POD checking
40% → ~0% manual

Invoice & receipt recognition — accountancy firms

The documents standard software couldn’t read — crumpled, folded, half-scanned — recognised by AI and booked with the right ledger and VAT codes, invisible to the user.

same pattern as: reading damaged CMRs
Instant answers

Customer e-mail AI — accountancy firm

Trained on internal data; answers routine client questions immediately and escalates to a human the moment it’s unsure or the tone escalates.

same pattern as: “where is my truck?” & creditor mails
What it’s really worth

The value is bigger than FTE savings

Saving a fleet manager or two is a welcome bonus — but the structural value sits in three places:

1 · Better fleet utilisation

Trucks stop losing hours to invisible waste: arriving at 06:00 at a customer that opens at 09:00, waiting unnoticed at a ramp, or getting the wrong follow-up trip.

3 wasted hours a day = a full driving day lost every week — per truck.

2 · Faster, proactive communication

Customers hear “tomorrow 10:00 instead of 08:00” from you before they ask. CS answers from live data instead of chasing fleet managers. Waiting costs come with evidence attached.

The “where is my truck?” mail answers itself — correctly.

3 · Your good people, bigger scope

Planners and fleet managers who think along get tools that remove the searching, recalculating and guessing — so the same people can confidently run more trucks, better.

One glance: ETA + remaining drive time, per truck. Decisions in seconds.
Straight answers

What we will and won’t promise

✓ What you can hold us to

  • A working automatic ETA in 12 weeks, judged on measured accuracy — not on slides.
  • Sprint 1 contains the ETA and nothing else — the rest of the demo is roadmap, not scope.
  • Everything documented and handover-ready from day one.
  • Read-only access during development; writes into Lovos only when proven and approved.
  • We flag it ourselves — early — if scope or capacity stops fitting.
  • A fixed person on the project who knows Vos processes and systems.

✗ What we won’t do

  • Start with a full AI planning system. It’s too complex to begin with — we build towards it in tested steps.
  • Sell a perfect ETA. It’s a window (±15–30 min), and we’ll show you exactly how accurate it is.
  • Park your product on our servers or behind our subscription.
  • Rebuild Lovos. Your IT owns the screens; we connect via API.
  • Promise everything at once. That movie has been seen before — finishing one thing well is the whole strategy.

One project. Finished. Working. Then the snowball.

To the ETA project team: the demo behind the button runs on fictional data, but every input in it exists in your systems today. Send us the concept and data overview your team already prepared, and within twelve weeks the automatic ETA runs on real Vos trips — finished, measured, and yours.

▶ Open the working ETA demo
Procai · Nijmegen · €125/h short engagements · €95/h multi-month projects · fixed price for fixed scope