Skip to content

Route optimisation

We build a model that schedules trips according to your rules, while your planner makes the decisions. For Fokker, we built a tool for planning work orders and capacity.

Fokker plans work orders, parts and capacity ahead of time using the planning tool we developed

The tool combines historical maintenance data, sensor data and operational data to align people, parts and hangar capacity. No routes, but the same question: how best to allocate scarce capacity.

Read the case study

How long does it take to plan tomorrow’s routes?

An experienced planner knows the customers and the roads, but as the number of addresses, time windows and vehicles grows, the puzzle becomes too complex to solve in your head. A model evaluates different options and presents the best one.

Book a call

Route planning that knows your rules, with the planner making the decisions

Discuss your planning

Rules as building blocks

Time windows, load capacity, driving and rest times, regular customers assigned to each driver, and vehicles that cannot access certain areas.

Collections and deliveries

Trips involving loading and unloading, return flows and multiple depots, all in one plan.

Replanning throughout the day

An urgent order or a traffic jam: the model recalculates the rest of the day and lets the planner choose.

Allocating capacity

Just as we coordinated people, parts and hangar capacity for Fokker, we allocate vehicles and drivers across orders.

Recalculate yesterday’s plan before the model plans tomorrow

  1. Rules and data

    We work with the planners to define the rules and gather orders, addresses and data from previous trips.

  2. Looking back

    The model plans a period for which the trips have already taken place, and we compare its plan with what actually happened.

  3. Alongside the planner

    First, it gives the planner a proposal. Then it is integrated into the planning software, using corrections to help it learn.

How complex does route planning need to be to warrant a model?

Assess it in a conversation

A good fit if

You plan routes to many addresses across multiple vehicles each day, with time windows or capacity limits, and the planning currently depends on one or two experienced planners.

Not a good fit if

You drive the same fixed routes every day, or only visit a handful of addresses. In that case, a standard route planner or a fixed list of trips is enough.

What determines the scope

The number of addresses and vehicles, how many rules apply, whether routes need to be replanned during the day, and which planning software the model needs to work with.

Does the model take over the planner’s job?

No. The planner knows about exceptions that aren’t recorded in any file. The model calculates and makes suggestions; the planner adjusts them and makes the decisions.

Plan based on experience, use a standard route planner or build your own optimisation model?

Plan based on experience

Choose this if you have few addresses and the planner can keep track of everything. The risk is relying on that one person.

A standard route planner

Choose this if your requirements are common ones: time windows, capacity and travel times. Specific agreements with customers often cannot be accommodated.

Your own optimisation model

Choose this if your requirements are unusual or the planning needs to work with other systems. Our advice: start with one depot.

We’ve built planning tools for mobility and maintenance before

For Fokker, we built a planning tool; for GVB, a web app that lets transport specialists communicate which stops are no longer in service. Data Science Lab has been around since 2016 and is a Microsoft Partner and Databricks Partner.

Book an introductory meeting

We’ll start by discussing how tomorrow’s schedule is put together today

Stef Stoelwinder

Stef Stoelwinder

Business Development

Stef is your point of contact. In our first conversation, we’ll discuss your trips, your rules and where your current scheduling runs into difficulties.

Which part of your route planning would you like us to run the numbers on first?

Tell us how many addresses and vehicles you plan for, and which software package you use. We’ll reply by email or arrange an initial conversation.

Enter a valid email address.

Let us know how we can help you.

Frequently asked questions

Route optimisation automatically determines which vehicle visits which addresses, and in what order, while meeting the requirements and minimising driving and waiting time. In the literature, this is called the vehicle routing problem.

You can read about more applications under AI applications.

Both. Planning itself is an optimisation problem solved by a solver. Machine learning helps with the inputs, for example by predicting driving and stop times from previous journeys.

Together, they produce a plan that works in practice.

Orders with addresses, quantities and time windows; vehicles and their capacities; drivers and their schedules; and driving times between addresses. Records of previous journeys help us estimate stop times accurately.

We start by looking at what you already have.

Often, yes, if the software can export orders and import a plan. The model then runs in the background, while the planner continues working in their usual interface.

We check this for each software package before building anything.

Yes. If an urgent order comes in, a vehicle breaks down or there is a delay, it recalculates the rest of the day, treating journeys already under way as fixed.

The planner sees the proposed changes and makes the decision.

Route optimisation deals with journeys and their sequence. For staff schedules, production or maintenance, read about optimising planning with AI.

The approach is similar: first the rules, then the model.

We agree in advance what matters: kilometres driven, on-time deliveries and vehicle utilisation. You measure these in a logistics dashboard, before and after implementation.

That way, you can see the difference in your own figures.

Yes. If you have a better idea of how many orders to expect in each region, you can schedule vehicles and drivers earlier. Read about forecasting and the mobility and logistics sector.

Forecasting and planning reinforce each other.