PE-backed trailer-fleet operator North America 11,000 trailers Live

A trailer fleet flags costly repairs as they happen, not at month-⁠end.

Before Mira
Deals were priced without asset-level visibility of maintenance cost, the cost that decides whether each lease makes money.
With Mira
Maintenance spend forecast for every trailer and checked against actuals.
98%

maintenance-spend forecast accuracy against actuals

11,000

trailers covered

Every trailer

forecast at asset level (maintenance spend per asset)

— Before Mira —

Deals priced without seeing the maintenance cost behind each trailer.

The operator leases trailers on maintenance contracts, so maintenance cost decides whether each deal makes money. Its deals were priced without asset-level visibility of that cost.

A forecast worth trusting was a project of its own: the last expert-built one took data scientists six months.

— Collect → Forecast → Alert —

One model, from work order to alert.

Collect

Work orders, parts and asset records across the fleet.

The records the fleet already keeps: what was repaired, which parts went in, and the details of each trailer. Mira ties every cost to the trailer it belongs to.

Forecast

Maintenance spend for every asset, checked against actuals.

Each trailer gets its own maintenance-spend forecast. Forecasts are back-tested against what was actually spent, so the team knows how far to trust them.

Alert

Repair-cost anomalies, flagged as they happen.

A repair that costs far more than the fleet norm is flagged as the work order comes in, not at month-end, so someone can check it while it still matters.

— The analysis —

How the forecast holds up.

Three views of one model, the way a buyer would check it. The charts use illustrative data; the live result is 98% maintenance-spend forecast accuracy against actuals.

Maintenance spend, forecast vs actual
Illustrative data
Fig. 1 In this illustration, a Q4 back-test made from September data stays within 3% of actuals, and Q1 is forecast at $4.2M against a $4.09M plan, $114K (about 3%) over.
Forecast error per trailer
Illustrative data
n = 1,000 trailers
Fig. 2 Forecast error for each trailer in an illustrative 1,000-trailer sample: most land within ±3% of actual spend.
Repair-cost anomalies
Illustrative data
8 weeks of sampled work orders
Fig. 3 Eight weeks of work orders from an illustrative sample of trailers. In this example, repairs above 2× the rolling fleet median are flagged as they happen.

Want a forecast like this for your fleet?

Book a call

— People stay in the loop —

Mira proposes.
The team approves.

The operator's team approves before anything happens, such as booking early inspections.

Proposal · North yard Illustrative data

North yard's 38 units are driving over half of the Q1 overrun (+$61K). Book early inspections for the worst 12?

Waiting for the team's approval

01 · Mira proposes

A next step, not just a number

Forecasts and anomalies arrive with a proposed next step.

02 · The team approves

Nothing happens without a yes

Each proposal waits for someone on the operator's team to approve it.

03 · Checked against actuals

Forecasts with a track record

Every forecast is checked against actual spend, so the team can see how far to trust it.

— What's live, what's next —

Running today, and what comes next.

Live

What runs today

  • Maintenance-spend forecasts for every trailer, at asset level.
  • 98% maintenance-spend forecast accuracy against actuals.
  • 11,000 trailers covered.
  • Repair-cost anomalies flagged as they happen.

Find out what your business might be missing.

We'll show you something you probably don't know about your business.

Next case study →

Case study 02 / 03 · Built Environment

A Saudi contractor prices a bid from evidence, not guesswork, in 15 minutes.

Book a call

Leave your work email and we'll be in touch to book it.

A short call

Tell us a bit more

Optional, but it helps us make the call useful.

A short call
success

Thanks. We'll be in touch

We'll email you to book your call.