The Empty-Mile Problem: The Fleet Cost Hiding Behind Healthy Utilization in 2026

The Fleet Cost That Doesn’t Show Up on the Utilization Dashboard

A fleet manager looking at a healthy utilization number can still be losing money on every route, because utilization does not distinguish between the miles that earn and the miles that do not. A vehicle can be out on the road all day, counted as fully utilized, and spend a large share of that day running empty: returning to the depot after its last drop, repositioning to another zone, or backtracking across a poorly balanced route. Those are empty miles, and they are the most expensive miles a fleet runs.

ATRI (American Transportation Research Institute) 2025 data puts deadhead (empty) at roughly 16.7% of all truck miles, about one mile in six runs with nothing in the back.

Empty miles cost exactly what loaded miles cost, fuel, driver time, vehicle wear, road charges, but they produce nothing. They are cost-per-mile with the revenue stripped out. And because most fleets measure utilization rather than the ratio of loaded to empty miles, the cost hides in plain sight: the dashboard says the fleet is busy, while a meaningful fraction of the busyness is a vehicle moving without a job to do.

This piece is about that hidden cost. It explains why utilization conceals empty running, where empty miles actually come from, and how dispatch and orchestration, rather than more vehicles, cut them. The empty mile is one of the few fleet costs that can be reduced without buying anything or cutting service, because it is not a capacity problem; it is a coordination problem, and coordination is what a modern dispatch system is for.

Why Utilization Hides the Empty Mile

Utilization, as most fleets measure it, asks whether an asset is deployed: is the vehicle out, is the driver working, is the shift being used. Those are useful questions, and the earlier fleet-utilization work covers them in depth. But they are questions about time and deployment, not about what the miles actually accomplished.

A vehicle can score well on every utilization measure and still run empty for a large part of its day. It was deployed, its shift was used, its capacity was booked, and yet much of its distance might have been covered with nothing in the back. Utilization counts the vehicle as productive because it was moving; it does not ask whether the movement was loaded or empty. That distinction is invisible to the metric and enormous to the cost.

Also Read: TMS: Decarbonizing European Supply Chain

This is why empty miles are so persistent. They do not show up as idle time, which fleets watch for, or as low utilization, which fleets manage against. They show up as a vehicle that looks busy, so nothing flags them, while they quietly inflate cost-per-mile across the whole operation. The first step to cutting them is measuring the right thing: not just whether the vehicle was used, but what share of its miles were loaded.

Where Empty Miles Come From

Empty miles are not random; they come from a handful of identifiable sources, and naming them is the start of removing them.

1. Return-to-Depot Legs

The most common empty mile is the drive home. A vehicle finishes its last delivery and returns to the depot empty, and the further that last drop is from base, the longer the empty leg. Multiplied across a fleet every day, the return-to-depot deadhead is often the single largest source of empty running.

Eurostat reports 21.6% of all EU road-freight vehicle-kilometres were run empty in 2024 — rising to nearly 26% for national (domestic) transport.

2. Repositioning

Vehicles are frequently moved from where they finished to where they are next needed, between zones, depots, or shifts, and these repositioning moves are empty by definition. Some repositioning is unavoidable, but much of it reflects a plan that did not consider where each vehicle would end up and what it would do next.

3. Unbalanced Routes

Poor territory design and weak sequencing create empty running inside otherwise loaded routes: backtracking, cross-hauling between overlapping territories, and criss-crossing that adds distance without adding stops. These miles are not fully empty, but they are unproductive, and they behave like empty miles on the cost line.

Also Read: TMS: Decarbonizing European Supply Chain

4. Missing Backhaul and Return Loads

When a vehicle returns empty, the question is whether there was anything it could have carried, a pickup, a return, a transfer, on the way back. Most operations have such work available but do not match it to the vehicles already heading that way, because the returns and the routes are planned by different systems, or not planned together at all.

5. Siloed Fleets

When owned vehicles, third-party carriers, and gig drivers are optimized separately, one fleet runs empty legs while another is over capacity nearby. The empty mile here is a coordination failure: the work to fill it existed, but in a different silo that the planning never crossed.

How Dispatch and Orchestration Cut Empty Miles

Every source of empty miles is a coordination problem, which is why the remedy is dispatch and orchestration rather than fleet size. Four capabilities do most of the work.

The first is backhaul and return-load matching: treating the return leg as capacity to be filled, and matching available pickups, returns, and transfers to the vehicles already heading that way. A return-to-depot deadhead becomes a productive leg when the system knows what work lies along it.

The second is balanced route and territory design. Sequencing routes and shaping territories to minimize backtracking and cross-hauling removes the unproductive miles buried inside loaded routes, and keeps last drops closer to where vehicles need to end up.

The third is dynamic re-optimization. Empty miles are often created mid-day, when a plan meets reality and a vehicle ends up somewhere unplanned. A system that re-optimizes as conditions change can redirect a vehicle that would otherwise run empty toward work it can pick up, rather than sending it home.

The fourth, and the largest for many operations, is multi-fleet orchestration. When owned, third-party, and gig fleets are planned as one system, the empty leg on one becomes the assignment for another. The work to fill empty miles usually exists somewhere in the network; orchestration is what lets the planning cross the silos to find it. Together, these turn the empty mile from an accepted cost of doing business into a target that shrinks every time the coordination improves.

Also Read: The Real-Time Routing Stack: How Big-Box Retailers Engineer Rapid Delivery at Scale

How This Works in Practice

Cutting empty miles requires a system that plans loaded and empty legs together, across every fleet, and re-plans as the day unfolds, which is what an agentic TMS does.

In Locus, the world’s first agentic Transportation Management System, a Dispatch agent and a Carrier agent plan routes and assignments across owned fleets, third-party carriers, and gig drivers as one system, optimizing against more than 250 real-world constraints inside a continuous Sense-Decide-Execute-Learn loop. That means return legs are matched to available pickups and returns, routes are sequenced to cut backtracking, and a vehicle that would otherwise deadhead can be redirected to work in another part of the network, including work that would have gone to a different fleet. Because it plans the fleets together rather than in silos, the empty leg on one becomes the loaded assignment for another.

Locus runs this across 1.5B+ deliveries for 360+ enterprise customers in 30+ countries at 99.99% uptime. In one anonymized deployment, a Fortune 50 enterprise running 4,500+ drivers raised its delivery execution rate from 75% to 92% through the same continuous, constraint-aware optimization that finds and fills empty running.

What This Means for a Head of Fleet

The practical shift is to stop trusting utilization as the headline efficiency number and start measuring the loaded-mile ratio, the share of miles that actually carried something. Utilization tells you the vehicle was busy; the loaded-mile ratio tells you whether the busyness was worth paying for, and the gap between the two is where the money leaks.

Freight inefficiency and empty miles cost US fleets an estimated ~$30 billion a year in fuel, labour, and underused capacity.

Also Read: AI Dispatch for Q-Commerce Rider Productivity in ID and PH

Empty miles are the highest-return efficiency target a fleet has, because unlike most savings they require no trade-off: cutting them does not reduce service, shrink capacity, or touch the vehicle mix. It recovers cost that was producing nothing. And because empty miles are a coordination problem, not a capacity one, the lever is the dispatch and orchestration system, not the size of the fleet. The question to ask is not whether the fleet is busy, but how much of its distance is empty, and what work is going unmatched to fill it.

Learn more, visit locus.sh.