Mid-Mile Planning Software for Hub-to-Hub Freight

Shyftbase Mid-Mile consolidates shipments moving between facilities — stores, FCs, DCs and last-mile sort centers. The system analyzes shipping patterns, consolidates loads and optimizes routes between those facilities, coordinating internal fleets and third-party carriers. Real-time tracking supplies status updates, ETAs and inventory movements to you, your contractors and your carriers.

Ask ten operators where the middle mile starts and you get ten answers, because most have never had to name it. It is the freight moving between facilities rather than to customers, and nobody at either end placed the order — which is also why the leg gets under-managed. It has no customer watching it, so it is measured late, by accountants.

Where the middle mile starts and ends

Shyftbase states the destinations it plans for in one line: shipments bound for stores, FCs, DCs and last mile sort centers. Read that as a boundary, not a brochure. A load leaving a distribution center for a retail store is middle mile. The same freight leaving that store on a van with a customer's name on the paperwork is not.

Three tests settle nearly every argument about the boundary:

• Who receives it. An internal facility, a franchise partner or a carrier's sort center means middle mile. The person who placed the order means last mile.
• What the handling unit is. The middle mile moves pallets, cages and trailers. The last mile moves cartons, parcels and appliances into a hallway.
• What breaks when it runs late. A late middle-mile load costs you a wave. A late last-mile stop costs you a customer.

The vocabulary is unhelpful, so state it plainly: mid mile, midmile and middle mile are one leg, and line haul is the older term for its trunk portion. None implies a different operating model, and a vendor treating them as separate products is selling a naming convention. If you came to understand the term rather than to buy anything, start on the self-serve plan.

The mid-mile warehouse is a clock

A mid-mile facility gets described by its floor plan and is run by its clock. Cross-dock, consolidation hub or sort center, its performance is decided by a short list of times: the inbound cut-off after which a trailer is worked tomorrow, the sort window, the outbound departure for each lane, and the receiving hours at the far end.

Those times are the constraint, and they are why consolidation plans fail in ways that look like transport problems and are not. Here is the failure with its cause. A store takes deliveries between 06:00 and 11:00. The hub's outbound for that lane departs at 21:40. A pallet clearing the sort at 22:10 waits a full day for the next departure, and because the store's window shuts at 11:00 that is a day and a half of delay, not an evening. Nothing about the truck, the driver or the road was wrong. The pallet landed on the wrong side of one number.

So a mid-mile warehouse is planned as a schedule rather than a space, and cross-docking freight — taking it from an inbound door to an outbound door without putting it away — is the default mode rather than an optimization applied to one. Storage in a mid-mile node is a symptom: an arrival and a departure did not line up, and the building is absorbing the difference.

What the Shyftbase mid-mile module does

Four things, and it is worth being exact about which four, because pages in this category tend to describe an ambition.

It analyzes shipping patterns, consolidates loads and optimizes routes between facilities — look at what moves between which nodes, group the shipments that can travel together, then sequence the resulting movements across the network rather than one lane at a time.

It dispatches across both internal fleets and third-party carriers. A consolidation only your own trucks can execute is a plan with a hole in it, because the lanes most worth consolidating are usually bought rather than run.

It tracks: status updates, ETAs and inventory movements for shipments between facilities, with visibility across distribution centers, stores and sorting facilities. One movement therefore appears both as a transport event and as a stock movement — the join most operators still do by spreadsheet on a Monday morning.

It gives contractors and third-party carriers visibility over their own orders instead of making them telephone for it.

Mid-mile runs on the same platform as first-mile and last-mile execution and connects to warehouse and carrier systems. Where freight is stored, counted and picked is the warehouse management module; where a completed movement becomes an invoice is automated freight billing, and what the third-party carrier that ran the leg is owed for it is carrier settlement. The images and documents the leg produces on the way — a label, a bill of lading, a photograph of a damaged pallet — are read by the AI module. IFS, Acumatica, Epicor and Sage Intacct are the named ERP and accounting connections.

Sizing a consolidation on one lane

A mid-mile business case turns on how many departures your fill rate is buying you, not on a vendor's percentage — and that is arithmetic you can run before anyone gives you a demo.

The model below is a calculation, not evidence: every row is a number to replace. One line carries a citation, and it is the line most likely to be wrong for you — the American Transportation Research Institute's 2025 industry-average operating cost of $2.336 per mile^[atri]. The sectors that release reports margins for are truckload, refrigerated, tank, LTL and flatbed^[atri] — a hub-to-hub trailer lane is closer to that mix than a delivery round is, and the report's own population is worth reading before borrowing the number for either.

What the model is sensitive to is the transferable part. The saving is not proportional to how much fuller the trailers get; it is proportional to whole departures deleted, because a trailer that does not run costs nothing. Fill moves in steps: on these inputs, 16 positions to 18 removes one departure, and 16 to 24 removes three. That is why a project can report a real gain in fill and no change in cost without anyone lying, and why the unit to plan is a corridor, not a lane.

The two constraints that decide whether a lane consolidates

Two hard limits sit under every consolidation plan, and neither is in the software's gift.

The first is weight, and it is not one number. On the Interstate System the maximum gross vehicle weight is 80,000 pounds, the maximum on any one axle is 20,000 pounds, and the maximum on tandem axles is 34,000 pounds — with two consecutive sets of tandems allowed 34,000 pounds each only where the first and last axle are at least 36 feet apart^[weight]. Loaders know the consequence and planning tools often miss it: a trailer can sit well under the gross limit and still be illegal because the weight is stacked too far forward. A consolidation that counts pallet positions and ignores where the mass lands passes on paper and gets turned at a scale house.

The second is the driver's day, and it is federal. A shift cannot open until the driver has had 10 consecutive hours off; once it opens, everything has to finish inside a 14-hour window, of which no more than 11 hours may be driving^[hos]. A consecutive 30-minute interruption falls due before more than 8 hours of driving time have passed, unless the driver qualifies for one of the short-haul exceptions^[hos]. Across a week the ceiling counts on-duty time rather than driving time: 60 hours in any 7 consecutive days, or 70 hours in any 8 where the carrier runs every day^[hos].

Together they give a decision rule to apply before designing anything for a lane. Take the round-trip driving time at your real average speed, add what the driver spends at both docks, and compare the total with the 11-hour driving limit and the 14-hour window^[hos]. Inside both, the lane is a turn: one driver, one day, the trailer back for the next sort. Past the driving limit but inside a legal day for two, it is a relay with a meet point. Past both it is an overnight, which changes the economics and not just the schedule, because the trailer leaves the pool for the next sort. Routing sequences the stops within a leg; what route optimization actually saves is a different question from whether the leg should run.

What carries over from first mile

First mile no longer has a page of its own here, so it is worth recording what was on that one: a network module for supplier communication, coverage of first mile, mid mile and last mile, and a reverse-logistics module running alongside the forward flow. It did not describe how inbound pickup, booking or dock scheduling work, so there is nothing on that subject to restate and nothing will be invented to fill the gap.

For anyone who arrived looking for first-mile software, inbound and outbound are the same planning problem with the arrow reversed. A collection from a supplier into a consolidation hub is a facility-to-facility movement with a time window at one end and a dock at the other, and it consolidates by the arithmetic above. The difference is whose clock you are on: outbound, the receiving window is yours to negotiate; inbound, it belongs to a supplier with no incentive to move it, which is why inbound consolidation starts as a commercial conversation and becomes a planning exercise only after that.

If your problem is genuinely inbound — purchase-order visibility, supplier booking, advance notice of what is on the truck — none of it is described in what this page inherits.

What this does not do

Four boundaries. A planning tool oversold is a planning tool switched off in month three, so these are worth reading before the business case rather than after it.

It does not create dock capacity. If the binding constraint is doors, shift length or people on the sort, pushing more freight through fewer trailers moves the queue rather than shortening it. Consolidation and cross-dock throughput are separate projects, and doing the first while the second binds buys a worse Monday.

The published constraint set is thin. What Shyftbase states is pattern analysis, load consolidation and inter-facility route optimization. Weight-and-axle legality, dock appointment windows and cut-off times are not named among them. Make that a demo question: what does the planner hold as a hard constraint, and what does it do with a load it cannot make feasible.

Visibility is a claim about a data source, and the source is not stated. Real-time status, ETAs and inventory movements are described; where those events come from — a driver app, telematics, a carrier's feed or a manual scan — is not, and neither is the refresh interval. Those answers decide whether an ETA is a measurement or an estimate of an estimate. The same gap covers the carrier side: whether a third-party carrier gets a portal, an app or a feed is unpublished, and it decides whether they use it or keep calling.

No figure here was measured on a Shyftbase customer. The arithmetic runs on assumptions you replace and one published industry cost per mile. There is no first-party outcome number here, because none exists that could be sourced — and an unsourced one is worth less than silence.