Most route planning stops at the depot door. A vehicle finishes its load, drives back to reload, and the next leg gets planned as a new route, disconnected from the one that just finished. Nobody decided that a driver's day should be built this way. It's how route optimization has drawn its boundaries for years, one trip, one plan, repeated as many times as the vehicle needs to return.
The two loops each look fine on their own. The gap between them is where the shift gets expensive.
A route plan that ends at the depot door only solves part of the day
When a vehicle needs to refill partway through a shift, most planning treats that return as the end of one problem and the start of another. The first loop gets optimized well. The second loop gets built afterward, sometimes by hand, without much knowledge of where the vehicle already is, how much of the shift is left, or what would sequence best given the first half of the day. Two loops planned apart are rarely as efficient as one day planned together, even when each loop looks fine on its own.
Idle time at the depot is time nobody bills for, but everyone pays for
A driver back at the depot for a reload is being paid whether the next load is ready or not. When the return trip wasn't planned as part of one continuous shift, that gap is common, the vehicle arrives before the next batch is staged, or before a bay is free, or before anyone has decided what the second loop should even contain. None of that time adds capacity. All of it adds cost. Among the costs that accumulate this way, paid idle time and the deadhead miles driving back to a depot that wasn't on the way to anything both inflate cost per delivery without moving a single extra order.
One driver's day, planned as two separate problems
Consider a single driver working a full shift with one scheduled reload.
- Morning, the first loop is planned well, tight sequencing, full capacity, finishing back at the depot by early afternoon.
- Midday, the vehicle returns to reload. Because this leg was planned as a separate trip, nobody staged the next batch to be ready at arrival. The driver waits twenty minutes for loading to catch up.
- Afternoon, the second loop starts as if the day just began, sequenced without any reference to where the driver already is or how many hours are left in the shift. A stop that would have been efficient earlier in the route now adds a longer drive back across the same ground covered that morning.
Each of those three moments holds up fine on its own. Stacked across one shift, they add up to real paid time with nothing to show for it, not because anyone made a mistake, but because three reasonable decisions were never connected to each other.
Depot returns are a planning decision, not a scheduling inconvenience
Treating a refill as a first-class part of route optimization, matching it to real depot operating hours, loading or charging duration, and allowed refill windows, closes that seam instead of working around it. The distinction matters more than it sounds. A depot return handled inside the same optimization as the rest of the shift can be sequenced, staged, and timed on purpose. A depot return handled as an afterthought is, by definition, something the plan reacts to rather than something it accounted for.

This isn't specific to one kind of operation
Heavy goods delivery hits this constantly, vehicles fill up fast and need multiple reloads across a shift. Grocery and other high-frequency delivery networks see the same pattern wherever a single vehicle can't carry a full day's orders at once. Multi-depot operations add another layer, since a return doesn't always mean the same depot the shift started at. Wherever a vehicle reloads more than once a day, the same shape tends to appear, a shift split into several disconnected trips carries more idle time and more empty miles than the same shift planned as one.
The takeaway
Count how many separate routes your planning tool creates for one driver in a single day. Every seam between them is a place idle time and empty miles can hide.

