Chuo Taxi — Japan
A Toyota Prius taxi at Chuo Taxi was measured across two applications over 106 days and 13,257 km. Fuel efficiency rose 11.5% after the first application and 43.2% after the second; across the whole trial the aggregate is 28.3%. This is a single vehicle — a strong signal, not a fleet result.
Urban taxi duty — constant stop-start, long idle periods, high annual mileage — is one of the hardest duty cycles for combustion efficiency, and one of the most sensitive to it.
| Operator | Chuo Taxi, Japan |
|---|---|
| Fleet context | 10 taxis — the cost projection in the published case study is based on the fleet, the measurement is not |
| Vehicle measured | 1 Toyota Prius (hybrid), odometer 225,942 km at start |
| Baseline | 12.224 km/L |
| Applications | Two: 15 Feb and 31 Mar 2017 |
| Trial period | 106 days, 13,257 km |
| Aggregate result | 15.68 km/L — 28.3% above baseline |
Every measurement, not just the best one
| Application | Measured on | Distance since (km) | Fuel efficiency (km/L) | Change vs baseline |
|---|---|---|---|---|
| 1st — 15 Feb 2017 | 26 Mar 2017 | 5,602 | 13.63 | +11.5% |
| 2nd — 31 Mar 2017 | 27 May 2017 | 8,655 | 17.51 | +43.2% |
| Whole trial | 106 days | 13,257 | 15.68 | +28.3% |
Method before result
The taxi ran normal urban service throughout. NanoEFX was applied twice, and fuel efficiency was measured after each application against the same 12.224 km/L baseline, with odometer readings recorded at each step.
As with the Toyama trial, the aggregate across the full trial — 13,257 km over 106 days — gives 15.68 km/L, or 28.3% above baseline. The 43.2% figure is the second measurement interval, not the trial result.
The limits of this trial
We publish these next to the result rather than in a footnote. If a limitation would change how you read the number, you should read it at the same time as the number.
Known limitations
- One vehicle. The published case study pairs the 43.2% figure with a ten-taxi fleet, which reads as a fleet result — it is not, and we say so here.
- A hybrid drivetrain complicates fuel-efficiency measurement considerably: the share of electric assist varies with traffic, battery state and driving, and none of that is documented for this trial.
- Urban taxi duty is highly variable — fares, routes, traffic and idle time change day to day. Over 106 days that variation is not controlled.
- There is no control group.
- A 43.2% improvement on any single vehicle should be treated as an outlier to be explained rather than a headline. We publish it because it is in the data, with the aggregate alongside it.
Reading this evidence in context
The honest reading is that this vehicle improved substantially and that we cannot fully explain why the second interval is so much larger than the first.
It is a reason to run a pilot on a taxi fleet with proper controls — not a number to put on a brochure.
Questions operators ask about this trial
Was this ten taxis or one?
Is 43% credible?
Why is the aggregate lower than the second measurement?
Provenance
- Operator
- Chuo Taxi, Japan
- Vehicle
- Toyota Prius (hybrid), odometer 225,942 km at trial start
- Sources
- Published case study dataset; Eco-Spray master test dataset (2020), anonymised entry matched by odometer and trial distance
- Trial period
- 15 February to 27 May 2017
- Last checked
- 15 August 2026
The only baseline that matters is yours
Every trial on this site was run on someone else’s vehicles, duty cycle and baseline. A NanoEFX pilot measures yours — and we will help you design it with a control group so the result stands up.