Quick answer: AI route planning takes your list of drops, delivery windows, vehicles and drivers, and works out an efficient order and route for each van far faster than planning by hand. For a small delivery business it can mean fewer miles, more reliable time slots and less stress each morning, provided your address data is clean and drivers trust the plans.
If you run a handful of vans, you probably still plan routes the traditional way: someone sits with a map, a spreadsheet or a sat nav and juggles addresses until the day looks workable. That works at low volume, but it breaks down as drops grow, customers ask for narrower time slots and a driver calls in sick. This guide explains what route optimisation software actually does, how to set it up and the UK rules that still apply to the people behind the wheel.
Table of Contents
What ai route planning does differently from a sat nav
A sat nav or mapping app answers one question: what is the best way from here to there? Route optimisation answers a harder one: given all of today’s stops and all of today’s vehicles, which van should visit which addresses, in what order, to meet every promise at the lowest overall effort?
Good ai route planning weighs several constraints at once:
- Time windows. A customer who is only in before noon, or a shop that only accepts deliveries before opening.
- Vehicle capacity. Weight, volume or number of crates, so no van is overloaded while another runs half empty.
- Driver shifts and breaks. Start and finish times, and when breaks need to happen.
- Service time. How long each stop takes, from a quick parcel drop to a two-person furniture carry.
- Traffic patterns. Typical congestion at different times of day, and in some tools live traffic.
- Priorities and skills. Urgent orders, or jobs only certain drivers are trained for.
The “AI” label covers a mix of optimisation algorithms and machine learning. Some tools learn from completed routes, adjusting how long particular stops or areas really take. The practical point is that software can test far more combinations than a person can in a morning.
Signs your business is ready
Route software is not essential for every business. A single van doing the same round each week may be well served by a mapping app. It becomes worth considering when you notice any of the following:
- Planning the next day’s routes takes more than a few minutes, or depends on one person who knows the patch.
- Customers complain about missed or vague delivery slots.
- Drivers regularly cross each other’s paths or return to base with deliveries still on board.
- You are adding vehicles, new postcodes or same-day orders.
- You want to offer customers live tracking and accurate arrival times.
If none of these apply yet, ai route planning can wait. Keep a simple record of planning time, mileage and late deliveries each week, so you have a baseline to compare against when the business grows and you do start a trial.
Seven steps to set it up
- Clean your address data. Check postcodes, add delivery notes such as “side gate” or “ring bell twice”, and fix duplicates. Bad addresses produce bad routes whatever the software.
- Record realistic service times. Time a week of typical stops so the plan reflects reality, not optimism.
- Set up vehicles and drivers. Enter capacities, shift patterns, start points and any restrictions, such as a van that cannot enter a low-bridge area.
- Connect your order source. Many tools import from spreadsheets, and some connect to e-commerce platforms or order systems directly.
- Run plans alongside your manual routes. For a week or two, compare mileage, finish times and missed windows before switching over.
- Give drivers the app. Most tools include a driver app with navigation, proof of delivery photos or signatures, and status updates.
- Review weekly. Look at late stops, failed deliveries and driver feedback, then adjust service times and constraints.
Well-known options include Circuit, OptimoRoute, Route4Me and the route features inside some fleet and courier management systems. Try two with your own data during their trial period rather than judging from the demo.
Worked example: a food wholesaler in Coventry
Imagine a small food wholesaler in Coventry with three vans supplying cafés, restaurants and corner shops across the city and surrounding towns. The operations manager plans routes each afternoon by hand. Most customers want deliveries before lunch service, and some kitchens can only take goods before a certain hour. Drivers often finish at very different times, and one van regularly returns with drops that could not be completed.
The business starts using ai route planning with a spreadsheet import from its order system. The manager spends the first week tidying addresses and adding notes about loading bays and access codes. She times typical drops and discovers that restaurant deliveries take much longer than corner shop drops because of stairs and cellar access; the software’s default service time was far too short.
After a fortnight of running software plans alongside her own, she switches over. Planning time shrinks, finish times become more even across the three drivers, and customers receive a text with an estimated arrival window. When a driver is off sick, she removes the van and re-plans, and the software redistributes drops across the other two, flagging the ones that cannot fit so she can call those customers early. The drivers’ local knowledge still matters: one points out that the plan sends him down a street where parking is impossible at that hour, and the manager adds that as a note so the route changes.
Costs, effort and getting drivers on board
Most route optimisation tools are charged monthly, often per driver or per vehicle, with free trials and sometimes free tiers for very small volumes. The main costs beyond the subscription are the set-up time to clean data and the first few weeks of adjustment. Fuel and time savings depend heavily on how efficient your current planning already is, so measure before and after rather than relying on vendor claims.
Driver buy-in makes or breaks adoption. Involve drivers early, explain that the aim is fairer, more predictable days rather than surveillance, and give them an easy way to flag problems with a route. Experienced drivers often know access issues and parking tricks that no algorithm will find on its own; building that knowledge into notes and constraints makes ai route planning better for everyone.
Route planning is one part of a wider picture that includes stock forecasting, warehouse picking and customer communications. Our overview of AI in logistics for small businesses covers those areas and how they connect.
Common mistakes to avoid
- Importing messy addresses and blaming the software for poor routes.
- Leaving default service times in place, so every plan runs late.
- Setting delivery windows tighter than you can actually meet.
- Switching everything over on a busy day with no trial period.
- Ignoring driver feedback about access, parking and loading.
- Using tracking data for purposes staff were never told about.
Legal and safety notes, plus a checklist
Software can plan a route that is legal on paper and unrealistic on the road. Build breaks and sensible shift lengths into your settings, and check which drivers’ hours rules apply to your vehicles. Many small vans fall under GB domestic rules rather than the rules that require tachographs, but that depends on vehicle weight and use. The government’s drivers’ hours guidance explains the different regimes. Drivers must not handle a phone while driving, so devices should be mounted and routes loaded before setting off.
Location tracking of drivers is personal data under UK GDPR. Tell staff what is collected, why, who can see it and how long it is kept, and avoid tracking outside working hours. The ICO has published guidance on monitoring workers that is worth reading before you switch tracking on. Customer names, addresses and phone numbers shared with a route tool also need a suitable contract and privacy notice.
- Addresses cleaned and delivery notes added.
- Realistic service times measured and entered.
- Vehicle capacities, shifts and breaks configured.
- Drivers’ hours rules for your vehicles checked.
- Staff told clearly about location tracking and data retention.
- Parallel trial completed before full switch-over.
- Weekly review of late and failed deliveries scheduled.
Key takeaways
- AI route planning allocates stops across vans and orders them to meet time windows, capacity and shift constraints.
- Clean addresses and realistic service times matter more than which tool you choose.
- Run software plans alongside manual ones before switching over.
- Involve drivers early and build their local knowledge into notes and settings.
- Respect drivers’ hours, phone laws and UK GDPR when tracking staff and handling customer data.
AI route planning: frequently asked questions
Is ai route planning worth it for just two or three vans?
It can be, particularly if you have time windows, varied stops each day or same-day orders. With fixed rounds that rarely change, a mapping app and a well-kept spreadsheet may be enough. A free trial using your real data is the quickest way to find out.
Can route software handle last-minute orders?
Many tools let you add or remove stops during the day and re-optimise the remaining routes. How smoothly this works depends on the tool and on how quickly drivers receive updates in their app.
Will drivers see this as being monitored?
They may, which is why transparency matters. Explain what data is collected and why, limit tracking to working hours and use it to fix problems rather than to blame individuals.
Does it work outside cities?
Yes. Rural rounds with long distances between drops can benefit a great deal, although narrow lanes, farm access and poor mobile signal mean driver notes and offline maps become especially important.
