Measured results
The VNPost Pilot in Ho Chi Minh City produced measured figures. The results page gives each one, what was measured and what the Pilot does not yet show.
Paid Pilot
A Pilot: route optimisation on your own data
A Pilot (a trial run) is a fixed eight week engagement on your real operating data, run in parallel: your dispatcher assigns orders as today, drivers install nothing. Acceptance criteria are signed before day one; it ends with a signed acceptance record and a decision to continue, stop or extend.
Before a paid Pilot
Two free steps come before a paid Pilot; the second uses the Pilot's own data template and validation.
See the live demo in the sandbox
Canned scenarios in the public sandbox. No upload, no signup.
A free one-day measurement of your data
One real operating day or shift of your own, against the plan you actually ran that day, using the same data template and the same validation the Pilot uses. It is also the data quality check: if one day is not usable, ten will not be either.
One-day measurementPaid Pilot
The fixed eight week engagement described on this page.
Contract
A decision to continue, once the signed acceptance record is in hand.
The eight week calendar
Two weeks of data preparation, six weeks of running. The running weeks start only once the data passes the quality gate.
- Weeks -2 to 0
Data preparation
Answer the questionnaire, fill in the template, validate until clean, sign the acceptance criteria.
- Week 1
Load and configure
Master data is loaded, rules are configured, the rule-conflict log is closed, and one test run is made.
- Weeks 2 to 3
Shadow runs
Ten operating days of shadow runs, with a daily rejected-records report.
- Week 4
Live shift
One live shift, odometer readings recorded, reconciled afterwards.
- Week 5
Report
The report is generated and reviewed together.
- Week 6
Acceptance
The acceptance record is signed and a decision made to continue, stop, or extend.
The data quality gateData preparation
The data quality gate: until it passes, the clock has not started
The six running weeks start once the data passes these eight checks.
- Points carry coordinates or a resolvable address
- Orders carry a usable weight
- Order site codes match the site list
- Timestamps are readable
- Baseline routes carry an identifiable vehicle
- Baseline route days carry per-stop weight
- The share of rejected records after one cleaning pass
- Peak orders per run
Each check's numeric bar is in the data template and the acceptance document.
Rejected records are always reported by reason and by day, never hidden in a footnote.
What the standard Pilot covers
The frame every Pilot shares; your vehicle count, stop count and days are fixed inside it.
| Fleet | 10 to 30 route vehicles, from one depot |
|---|---|
| Service points | 30 to 150 points |
| Per-run size | At most 300 orders and at most 200 vehicles in one optimisation run |
| Historical data | At least 10 operating days, including one Monday, one Friday, and your known peak day |
| Order cut-off times | 1 to 3 a day, fixed from the start |
| Cargo | One cargo group, general cargo |
| Field verification | A shadow run, plus one mandatory live shift with odometer readings |
| Hosting | Hosted on SmartWay's cloud, a separate tenant per customer |
Six acceptance criteria
Each threshold, and what happens if it is missed, is signed before day one, so nothing is negotiated after the numbers are known.
- Reduction in kilometres per tonne of cargo, against your own baseline
- Share of supplied orders successfully routed
- On-time-window rate in the plan, not lower than the baseline
- Per-leg load breaches
- Live shift: odometer against plan
- Rejected-record rate
A missed criterion takes one of three agreed paths: extend, narrow the scope, or stop.
How we measure
Every figure published from a Pilot follows the method and rules below.
Three comparison lanes
- Baseline plan: the plan you run today, before Route4Green.
- Optimised plan: the plan Route4Green produces on the same data.
- Actually driven: distance and time recorded during the live shift.
Each figure names the two lanes it compares. The default denominator is tonnes of cargo, with the conversion factor printed in the table. Baseline km are labelled planned or odometer. Any field claim needs at least one live shift with real trucks and odometer readings.
Honesty rules for every published figure
- Every percentage carries its denominator and its conversion factor.
- We never present kilograms per route as a measure of how full the vehicles were, or as capacity utilisation.
- Working hours are only compared when both sides use the same time model.
- Cost is compared only when both sides include the same components; otherwise it is not published.
- Any emissions figure would use volume-normalised kilometres and a sourced factor printed beside it. None is published today.
- No feasibility claim is made without a per-leg peak-load check.
- A metric the supplied data cannot support is reported as not measured, never estimated.
What a Pilot does not conclude
- Results come from limited operating data and promise nothing about ongoing live operation.
- No commitment on availability, response time or out-of-hours support.
- Results for one kind of transport are not extrapolated to another.
- A 2 to 4 week measurement is not a service commitment.
When we decline a Pilot
- No like-for-like baseline: no per-stop weight and distance, and no fixed routes serving only the in-scope points.
- Peak orders per run exceed the technical cap and the scope cannot be split.
- No named operations owner, or the customer will not sign acceptance criteria up front.
.xlsx
Pilot data template
Four sheets with fixed column names: orders, vehicles, sites, baseline_routes. The per-stop weight column is what makes the raw figures publishable.
Frequently asked questions
Does a Pilot change how we dispatch today?
No. Route4Green runs as a shadow run; your dispatcher keeps assigning orders the way they do today throughout the Pilot.
What data do we need?
The four sheets in the data template: orders, vehicles, sites, baseline_routes, with at least 10 operating days.
Who does the work on our side?
One operations owner who signs the acceptance criteria, and one data contact who answers within a working day.
What happens if a criterion is missed?
The three paths agreed from the start: extend, narrow the scope, or stop.
Is there a fee?
Yes. The Pilot is paid and quoted by scope. No figure is published on this page.
A free one-day measurement of your data
One real operating day or shift of your own, against the plan you actually ran that day, using the same data template and the same validation the Pilot uses. It is also the data quality check: if one day is not usable, ten will not be either.