Your dispatcher's plan works. Here's what it's costing you to leave the gap open.

Here is what he cannot do, and what nobody can do: check every pairing.

Why the gap exists, and why it is not his fault

Five drivers. Fifteen loads, three each, in every order. The number of ways to arrange that board is 1.3 trillion. A real fleet — fifty, seventy-five, a hundred trucks against hundreds of loads and dozens of driver clocks — is larger still.

So the plan he builds is good. It works. Trucks roll, loads move, customers get served. What it cannot be, built by hand between 3 and 7, is the best one. And that gap is not a reflection of his skill. It is a reflection of arithmetic. No person alive can compare 1.3 trillion arrangements before the shift starts.

That gap has a cost. It shows up as deadhead miles on lanes that sat next to a pickup nobody caught. It shows up as two drivers running hard while two others sit light. It shows up as an HOS problem found after the plan went out.

What inaction costs, in numbers you can reproduce

Deadhead miles are not just wasted time. They are diesel you burned for nothing and parts you wore out for nothing.

At a conservative 6% mileage reduction — below the 12–16% Nivio typically delivers — a single truck runs roughly 3,600 fewer empty miles per year. At $5.00 per gallon diesel and $0.20 per mile in maintenance, that is $4,720 per truck per year in fuel and upkeep you never spend. On a 75-truck fleet, that is $354,000 a year. On 120 trucks, $566,400. These figures are modelled from the algorithm against representative bulk-fleet data — not customer results. The model is conservative by design and reproducible: run it yourself at niviolab.com.

The real customer numbers are larger. On a 75-truck sand fleet, Nivio moved loaded-mile percentage from 43.25% to 47.84%. Measured impact: $599,000 a year. On a 120-truck fuel fleet, the shift ran from 46.92% to 52.79% loaded. Measured impact: $1.48 million a year. Both of those fleets were already better-run than average when they started. They still found that much.

Every week you build the plan by hand is a week the gap stays open.

What the best plan actually looks like

Nivio's Proprietary Algorithm is a constraint solver. It searches the combinations — every load, every driver — against over 50 rules: hours of service, home windows, commodity type, weight limits, permit requirements, customer time windows. It returns the plan with the fewest empty miles in under 5 minutes.

Three things change on the board immediately:

1. Loads chain. One drop sits near the next pickup, so deadhead becomes paid freight instead of a cost centre.
2. Every load is legal before he looks. HOS and driver windows are checked for every assignment before the plan reaches him. He is not catching problems after the fact.
3. Work spreads evenly. The day distributes across drivers instead of running the same two hard while others sit light. Fewer complaints. Lower turnover pressure.

Planning time drops from 2–3 hours per day to under 30 minutes. The solver does the arithmetic. The dispatcher does the deciding — locking loads, moving assignments, calling the exceptions only he knows to call. That is not a reduction in his role. It is a return of his afternoon to actual dispatch work.

The cost of one more week

The modelled $4,720 per truck per year works out to roughly $90 per truck per week. On a 50-truck fleet, that is $4,500 sitting in the gap between the plan that works and the best one — every single week. That figure is modelled, conservative by design, and reproducible on the website in one click.

The plan your dispatcher builds is good. It has always been good. The question is how much the gap between good and best is worth leaving open.

Take the five-driver challenge at niviolab.com. Build the board yourself — five drivers, fifteen loads. Then hand the same scenario to Nivio and compare the miles. The math will be there when you are.

Every load, every driver, planned in minutes — on top of the TMS you already run. Works beside McLeod, TMW, Samsara, and Motive deployments in week one. Nothing gets replaced.