Planning Tomorrow's Drivers and Routes Before Five O'Clock

At a regional carrier running fifteen trucks, tomorrow's orders arrive all afternoon: emails with PDF delivery notes, calls from regular customers, bookings through two customer portals and a spreadsheet from a distributor. At 6pm, the dispatcher sits down with a whiteboard and a spreadsheet and starts building the plan. He knows which trucks have tail lifts, which drivers are close to their hours limit, which customers insist on morning delivery and which loading dock always runs slow. By 8pm the plan is done and texted to drivers. At 5:40 the next morning, a driver calls in sick and the plan has to be rebuilt from the cab of the dispatcher's car.

Dispatch in a small transport business works because experienced dispatchers carry an enormous amount of knowledge in their heads, which is also the weakness of the arrangement. The plan takes hours to build, it depends on one or two people, it can't easily be rebuilt when something changes, and the routes are only as efficient as a tired person can make them at the end of a long day.

This article covers how to get the constraints out of the dispatcher's head and into a system, and what automated planning and replanning look like around a transport management system and driver app.

What evening planning costs

The dispatcher's time is the obvious cost, along with the risk of depending on one person who can't take a vacation without the operation suffering. Routes planned by hand under time pressure tend to leave miles on the table, and every unnecessary mile costs fuel, driver time and vehicle wear.

Missed delivery windows and late plans cost too. Drivers who receive their run at 8pm or 6am can't prepare, customers can't be told their window until the plan exists, and every late change ripples through the day.

Write the constraints down first

Set an order cut-off and one intake point

Agree a cut-off time for next-day orders, and get every order into one system, whether that's your transport management system or a shared order board, with the information planning needs: verified address, delivery window, pallets or weight, and special requirements. Late orders can still be accepted, as long as they're clearly marked as late so everyone can see what was squeezed in.

Record vehicle and driver constraints

For each vehicle, record capacity, equipment (tail lift, refrigeration, truck-mounted forklift), and any restrictions. For each driver, record license categories and endorsements, site inductions, and availability, including their remaining hours under the drivers' hours rules that apply to you, such as hours-of-service in the US or tachograph rules in the UK and EU.

Capture customer and site knowledge

Write down what the dispatcher knows about each regular delivery point: opening hours, booking requirements, access restrictions for large vehicles, and typical time on site. That slow loading dock should appear in the system as a longer service time, so the plan accounts for it.

Agree priorities and publish time

Decide how to prioritize when not everything fits: by customer, by service level, or by order date. Set a time by which the plan is published to drivers, such as 5pm, and a process for changes after that.

What automated planning does

1. Collects orders. Orders arriving by email or as PDF delivery notes are read and entered into the order system automatically, with addresses checked and any missing information flagged for the office to chase. Portal and spreadsheet orders are imported directly.
2. Proposes the plan. At cut-off, the system builds a proposed plan: which loads go on which vehicle, which driver takes it, and the order of stops, respecting capacity, equipment, delivery windows, site restrictions and drivers' available hours, while keeping total miles and time down.
3. Shows what doesn't fit. Orders that can't be planned within the constraints are listed with the reason, such as no vehicle with a tail lift available or a window that conflicts with another drop, so the dispatcher can make the call.
4. Dispatcher adjusts and approves. The dispatcher reviews the plan, moves loads where their judgment says so, and approves it. The system checks that changes don't break any constraint.
5. Publishes to drivers and customers. Drivers receive their run in their app or by text, with stop details, site notes and contacts. Customers can be told their delivery window the same evening.
6. Replans when things change. When a driver calls in sick, a truck breaks down or an urgent order arrives, the system proposes options for reassigning the affected work, showing the impact on other runs, and the dispatcher picks one. Affected drivers and customers are updated.
7. Learns from the day. Actual arrival and departure times from the driver app are compared with the plan, so service times and drive-time estimates improve with each week.

The AI work is in reading messy inputs, such as emailed orders and PDF delivery notes, and in explaining the plan and its trade-offs in plain language. The plan itself comes from route optimization working within the rules you've written down, which keeps it predictable and auditable.

Our staff scheduling and dispatch system handles this for transport and field service businesses, and pairs with proactive client updates so customers hear about their window without calling. The logistics and transport page shows how dispatch fits with invoicing and compliance tracking.

A worked example

Picture the fifteen-truck carrier from the opening. Over two weeks, the dispatcher writes down the rules he applies, and the list grows to dozens of constraints: customers that only accept deliveries before 11am, a retail park with a height restriction, drivers without the license category for the largest vehicle, a distribution center where unloading always takes an hour. The company sets a 3pm cut-off for next-day orders. Emailed orders are read and entered automatically, with two incomplete addresses flagged for the office to chase. At 3:15, the system proposes a plan that covers all but four orders, explaining that those need a tail-lift vehicle that's already full. The dispatcher moves one order to the next morning with the customer's agreement and assigns the others to a hired vehicle. The plan reaches drivers at 4:30. The next morning, when a driver calls in sick, the system proposes two ways to cover his run, and the dispatcher picks one before 6am.

What to watch out for

• Unwritten knowledge. The first plans will miss things the dispatcher knows but nobody wrote down. Treat each of those as a constraint to add, and expect a few weeks of this before the plans feel right.
• Plans with no slack. A plan that's perfectly efficient on paper falls apart at the first delay. Build in buffer time, especially around busy sites and the end of drivers' shifts.
• Drivers' hours. Planning must respect the hours rules that apply to your drivers. Use their actual remaining hours where your systems provide them, and never let a plan depend on breaking a rule.
• Fairness to drivers. Drivers notice when the hardest runs always land on the same people. Balance difficult runs across the team, and explain how plans are made.
• Customer priorities. An optimizer will treat all deliveries as equal unless told otherwise. Make sure your priority rules reflect which customers and service levels matter most.
• Address and window quality. Bad addresses and missing windows produce bad plans. Fix the order intake before blaming the optimizer.

How to tell whether it's working

Track the time the plan reaches drivers each day and the hours spent planning. Measure miles per delivery and on-time delivery against promised windows. Count how often the dispatcher changes the proposed plan and why, because each change is either a missing constraint to add or a genuine judgment call. Track how quickly replanning happens when something goes wrong in the morning. And check drivers' hours data to make sure plans stay comfortably within the rules.

Where to start this week

Ask your dispatcher to write down, over one week, every rule they apply while planning: vehicles that can't go to certain sites, customers who need morning delivery, drivers who can't do certain runs, docks that always take longer. The list is usually longer than anyone expects, and it's the most valuable document in your planning process. At the same time, measure how long planning takes each day and when the plan reaches drivers.

Then set an order cut-off and move every order into one place. With the constraints written down and orders in one system, a proposed plan at cut-off and quick replanning in the morning become realistic, and your dispatcher gets to go home before the evening is gone.

The free PATH report gives you a quick read on where dispatch sits among the other jobs worth automating in your operation. To talk through how you plan today, book a call.