AI route optimisation: from VRP solver to live dispatch

Last-mile delivery, field service, waste collection or courier networks: every route plan is at heart a Vehicle Routing Problem. We build custom AI route optimisation software that solves VRPTW, CVRP, dial-a-ride and multi-depot scenarios with OR-Tools, VROOM and Hexaly, fed by real-time traffic data from HERE, TomTom or Mapbox. The result: shorter routes, fewer empty kilometres, better ETAs and a dispatcher with time left over for exceptions.

VRPTW & CVRP solver Real-time re-routing Dynamic dispatching ETA prediction Last-mile clustering
Discuss your routing case View use cases
Depot Eindhalte

What AI route optimisation is and isn't

Not every planning tool with a map attached is route optimisation. We build solutions that explicitly model the underlying combinatorial problem, and that problem has a name: the Vehicle Routing Problem.

Ever since Dantzig and Ramser formulated it in 1959, the Vehicle Routing Problem (VRP) has been one of the most studied optimisation problems in operations research. In practice it appears in dozens of variants. The Capacitated VRP (CVRP) accounts for the load capacity of each vehicle, the VRP with Time Windows (VRPTW) for customer arrival windows, the Multi-Depot VRP for multiple departure bases, the Pickup-and-Delivery Problem (PDP) for linked pickup and drop-off addresses, and the Dial-a-Ride Problem (DARP) for passengers who must be picked up and dropped off within travel-time quality standards. Each of them calls for a different solution strategy.

A VRP with twenty stops and one vehicle can be solved on a laptop. A daily plan for a fleet of sixty delivery vans with time windows, capacity limits, driving and rest time regulations and stochastic traffic information cannot. That is what solvers are for: open-source libraries such as OR-Tools and VROOM, or commercial engines such as Hexaly (formerly Localsolver). The trick is not in running such a solver, but in correctly modelling your operational reality and integrating the output into your existing planning and on-board computer stack.

"AI" appears in several places in this landscape. It is in the metaheuristics (simulated annealing, tabu search, Large Neighbourhood Search, genetic algorithms, Lin-Kernighan variants) that solvers use to find good solutions in limited time for problems where exact methods are too slow. It is in machine-learned ETA prediction based on historical journey times and real-time traffic. And it is in dynamic dispatching, where reinforcement learning or online optimisation decides how best to assign an incoming order to a moving vehicle without throwing the whole daily plan into disarray.

VRP variants we model

Choosing the right solver starts with the right problem formulation. Below are the variants that come up in practice in Dutch fleet and logistics operations.

🕒

VRPTW: time windows

Customers expect delivery between 09:00 and 12:00, or an engineer between 13:00 and 17:00. The VRP with Time Windows adds an earliest and latest arrival time to every stop, plus optional on-site service time. This is crucial for next-day delivery, field service planning and healthcare logistics. We model both strict windows and soft windows with penalty costs for minor overruns.

📦

CVRP: capacity constraints

A refuse lorry has a load volume, a van has a weight limit, a courier has a number of parcel slots. Capacitated VRP ensures no vehicle exceeds its maximum and plans intermediate unloading or loading runs at depots or transfer points. It can be combined with multiple compartments (multi-compartment) for chilled plus ambient goods, or residual waste plus plastics, metal and drink cartons.

🔁

PDP & DARP

Pickup-and-Delivery links a pickup address to a drop-off address: the vehicle that collects here delivers there. DARP adds passengers, such as WMO transport (Dutch municipal mobility support) or school transport with a maximum travel time per person. Both require sequencing constraints (load first, then unload) and ride-time limits, which makes the search space harder than a standard VRP.

🏭

Multi-depot VRP

Does your organisation have multiple bases, cross-docks or microhubs in the city? The Multi-Depot VRP assigns the best starting depot to each route based on the geographical spread of customers and the available capacity at each location. Essential for national courier networks and urban last-mile distribution with zero-emission zones.

⏱️

VRP with breaks and driving and rest times

European driving and rest time legislation (Regulation 561/2006) and the Dutch Working Hours Decree provisions for delivery services and couriers require breaks, daily rest and weekly rest. Our solvers include mandatory breaks explicitly, with flexible time windows and compatibility with digital tachograph data. No more schedules that look right on paper but result in a violation in practice.

🔄

Dynamic VRP

Same-day delivery, taxi dispatch and emergency repairs don't wait until tomorrow. Dynamic VRP decides online: a new order comes in, which vehicle picks it up, and how does that fit into the current route without harming the SLAs for other stops? We build insertion heuristics and rolling-horizon rescheduling that re-evaluate the state of the world every few minutes.

How Appfront builds route optimisation

We don't build a black-box app where you simply shout "stops in, route out". Our approach starts with your operational reality: which vehicle types, which contractual SLAs, which depots, which delivery arrangements, which exceptions your planners currently resolve by hand. Only once those are explicit do we write the objective function and the constraints the solver uses.

In practice, we build three layers on top of each other. At the bottom is the routing engine: for road distances and duration matrices we use OSRM, Valhalla or GraphHopper on your own infrastructure if the data allows, or HERE Routing, TomTom Routing and Mapbox Directions when you want ready-made traffic data and truck attributes. On top of that runs the solver: OR-Tools for most classic VRPTW/CVRP cases, VROOM for fast constructive heuristics on hundreds of stops, and Hexaly for the hardest problems where mixed objectives and complex constraints conflict. On top of that sits a dispatch layer that streams to your on-board computer or driver app and tracks what actually happens.

We work iteratively, with measurable results at each phase. A proof of concept on a week of historical orders shows in figures how many kilometres and hours the solver saves compared with your current planning. Only when that is convincing do we move on to production integration with your TMS, ERP and on-board computer.

From first planning question to production rollout

Routing projects stall when the solver is finished before the business logic has crystallised. Our phasing tackles this in the right order.

Constraint discovery

Together with your planners and drivers, we map out which hard and soft constraints actually apply: time windows, vehicle capacities, depot opening hours, customer-specific restrictions, breaks and driving and rest times. The output is a formal problem definition the solver can work with.

POC on historical orders

Using an extract of your historical trip and order data, we run the first solver runs in OR-Tools or VROOM. We compare the output with your actual routes from the same week: kilometres, hours, vehicle utilisation. Without that comparison, every routing figure is just marketing.

Integration and dispatch loop

The solver is connected to your order source, mapping service and on-board computer. We build a dispatch loop that calls the right optimisation for every new planning session or incoming order and sends the result to the driver app or on-board computer.

Live tuning and monitoring

A production routing system needs monitoring: why does the solver choose this suboptimal trip, why is the ETA error on corridor X larger than elsewhere? We instrument the pipeline, retrain ETA models on fresh trip times and adjust objective weights as the market changes.

Technology behind the routing stack

The choice of solver, routing engine and data sources depends on your scale, geography and compliance requirements. We have no preferred vendor that is "always the answer"; we choose based on what fits your problem.

For the optimisation itself, Google OR-Tools is our first workhorse. It is open source, production-ready, supports VRPTW, CVRP, PDP, multi-depot and breaks via the Routing Library, and has Python, C++ and Java bindings. For very large constructive runs on thousands of stops, we turn to VROOM, which is written purely in C++ and designed for speed. For cases where the objective function is heavily mixed, for example minimising kilometres while also balancing workload and respecting customer priorities, Hexaly (Localsolver) pulls ahead where OR-Tools struggles.

For road distance and duration matrices, we integrate with multiple services depending on the case. HERE Maps is strong on truck attributes: permitted bridges, height profiles, hazardous goods routes. TomTom provides accurate European traffic data and historical speed profiles. Mapbox Directions is a good choice for consumer-oriented apps with attractive map rendering. For self-hosted scenarios with large query volumes, we run OSRM, Valhalla or GraphHopper on your own infrastructure, with OpenStreetMap as the base data. For public transport and GTFS-RT feeds, we connect directly to NDOV-loket for Dutch public transport data when multimodal routing is needed.

OR-Tools VROOM Hexaly Lin-Kernighan Simulated annealing Tabu search Genetic algorithms HERE Routing TomTom Routing Mapbox Directions OSRM Valhalla GraphHopper GTFS-RT Python FastAPI PostGIS Kafka
Not sure about a large project yet?

Test your idea first: a working prototype in 1 day

With OneDayBuild, we make your idea tangible in a single day for €1,150, so you know whether further development is worth the investment. Decide to go ahead with the full build? Then we deduct the cost in full.

View OneDayBuild →

Applications by sector

Routing software has a different objective function and a different story in every sector. A refuse collection lorry optimises differently from a bicycle courier.

Last-mile and e-commerce

Parcel carriers with growing volumes and shrinking delivery margins need a dynamic VRP that respects customers' delivery windows, schedules failed deliveries for a second attempt and avoids zero-emission zones in cities for diesel vans. With postcode-level clustering and re-optimisation during the trip, our clients get more stops per hour from the same fleet, without making drivers drive harder.

Refuse collection and waste management

Municipal and commercial waste collection services plan weekly routes for residual waste, organic waste, paper and plastics/packaging streams, often with containers that need emptying on demand. CVRP with multi-compartment vehicles, sensor-driven fill-level data from smart bins and hard time windows for disposal sites deliver a plan that minimises driving and avoids empty trips to overfull containers.

Courier services: same-day and next-day

For next-day couriers, everything revolves around start-of-day optimisation: distributing all orders across drivers and vehicles in a single batch. Same-day couriers need a dynamic VRP that continuously slots new orders into routes already under way. For urban bicycle couriers we also integrate cycling routes and elevation differences, because Mapbox car routing has no suitable answer for that.

Field service and mobile workforce

Engineers for lifts, climate systems, telecoms or central heating boilers work with appointment windows, required skills per job and spare-parts stock in the van. A routing engine plus skills matching schedules the day so the right engineer arrives at the right time with the right spare parts. ETA updates to the customer via SMS or app happen automatically.

Dial-a-ride: WMO, student and taxi transport

Targeted-group transport and classic taxi dispatch are DARP problems: multiple passengers per vehicle, individual pick-up and drop-off windows and hard limits on travel time per person. We build solutions that allow ride-sharing where that is legally and qualitatively permitted, and that keep strictly separate rides where regulations require it, such as medical taxis.

Construction logistics and home-delivered meals

Construction logistics with cross-dock hubs and tight just-in-time slots on site calls for multi-depot VRP plus appointment scheduling. Restaurant delivery demands ultra-short cycle times and near-real-time rescheduling. What both have in common is that ETA prediction based on traffic data is crucial for predictable delivery.

ETA prediction and real-time re-routing

A plan is a snapshot. Traffic, weather, customer delays and no-shows mean reality departs from it the moment the first driver pulls away. That is where AI route optimisation takes the second step.

Machine-learned ETA

Standard map services give a traffic-average ETA. We train models on your own historical journey times, factored by driver, vehicle type, time of day and customer stop time. The result: ETAs that reflect your reality, not that of an average TomTom user. This matters for SLA reporting, customer notifications and realistic capacity planning.

Real-time re-routing

In the event of a traffic jam, an accident or a longer-than-expected stop, the system recalculates the remaining route for the vehicle concerned and, where necessary, passes stops on to colleague drivers. Re-optimisation takes seconds using insertion heuristics, rather than re-running the entire day's plan.

Geofencing and stop detection

Geofences around depots and customer addresses automatically detect arrival and departure, without the driver having to press a button. The data feeds the ETA models, triggers customer notifications and provides evidence for proof of delivery. For restricted zones (ports, industrial sites), geofencing also supports automatic check-ins.

Live traffic feeds

HERE Traffic, TomTom Traffic and GTFS-RT for public transport deliver speed updates per road segment in real time. Our dispatch layer feeds that data back to the routing engine so re-optimisations run on current speeds. For critical corridors we monitor incident feeds proactively and flag routes that pass through problem areas.

Why Appfront for route optimisation

OR engineers, not plug-in installers

Our routing projects are led by engineers who can instrument OR-Tools in Python and C++, write callbacks for specific constraints, and know the difference between first-improvement and best-improvement in local search. That difference decides whether your planning takes ten seconds or ten minutes.

Transport and logistics domain knowledge

We have been building transport and logistics software for years. We know the difference between a digital tachograph and an on-board computer, between eCMR and CMR, between a TMS and a fleet management app. That domain knowledge translates directly into better modelling of your VRP.

Integration-first mindset

A routing engine that doesn't talk to your order source, ERP and on-board computer is an academic exercise. We build integrations with Adaption, transport planning APIs, fleet tracking dashboards and on-board computer systems, so the optimisation lands in the system your drivers already use today.

Frequently asked questions about AI route optimisation

What is the difference between route planning and route optimisation?
Route planning determines the best order of stops for a single vehicle; in essence, that is a Travelling Salesman Problem. Route optimisation assigns stops to multiple vehicles at once, with capacity limits, time windows and breaks, which makes it a Vehicle Routing Problem. The combinatorial space of a VRP grows much faster than that of a TSP, so you need heuristic solvers such as OR-Tools, VROOM or Hexaly to find a workable solution within seconds or minutes.
Do you work with OR-Tools or with a commercial solver?
Both. For most classic VRPTW and CVRP cases, Google OR-Tools solves the problem perfectly well, and because it is open source you are not tied to a per-vehicle or per-call licensing model. For very large or very complex objective functions, such as multi-criteria optimisation, hard plus soft constraints and non-convex objectives, we use Hexaly because its solution times are significantly better there. We always document why we make a given choice.
Which map service is best for truck routing?
For truck routing with truck attributes (bridges, weight restrictions, height restrictions, hazardous goods routes), HERE is generally the strongest choice, with TomTom as a solid European alternative. Mapbox Directions supports truck routing but less extensively for Europe. For self-hosted scenarios, Valhalla or GraphHopper runs with a truck profile on OpenStreetMap data, which works well for mid-sized fleets if you want to be independent of API credits.
How do you handle driving and rest times?
We model breaks explicitly as constraints in the solver. For heavy transport (over 3.5 tonnes) we follow Regulation 561/2006: 4.5 hours of driving, then a 45-minute break, daily rest of nine or eleven hours, weekly rest of 24 or 45 hours. For lighter transport and couriers we follow the Dutch Working Hours Decree (Arbeidstijdenbesluit). The solver may only schedule a trip if the driver stays compliant, and the planning can be audited against tachograph data.
How accurate are the ETAs?
Standard map ETAs typically have an error of 10–20% on urban trips. With machine-learned ETA models that we train on your historical trip times, including stop times, driver effects and customer-specific factors, we bring that error back to 5–10%. More importantly, the spread of the error becomes more predictable, which makes SLA reporting reliable. We don't make blanket promises about this without data from your own operation.
Does your solution work for same-day delivery?
Yes. Same-day delivery calls for a dynamic VRP, because orders come in while vehicles are already on the road. We implement rolling-horizon optimisation: every few minutes the current situation is reassessed and new orders are inserted into routes already under way using insertion heuristics. Recalculating the entire day's plan is rarely necessary, and it would only confuse your drivers.
Can you integrate with an existing TMS or on-board computer?
Yes, that is usually the norm. We build integrations with TMS systems such as Adaption, with on-board computer suppliers and with fleet-tracking platforms. The solver receives orders and vehicle information via an API and sends routes back to the on-board computer or driver app. We avoid parallel systems wherever possible, so your existing TMS remains the source of truth for planning status.
What data do you need for a proof of concept?
Around two weeks of historical order data with customer addresses, requested time windows, volumes and weights; your vehicle deployment over the same period (which vehicle drove which route); and ideally the actual arrival and departure times per stop. With that input we run our solver and compare the result with your actual routes. Only then will you know whether the savings are large enough to continue.

Want to make your routing challenge concrete?

Send us your planning question. A week of historical orders is enough for a first solver run. No obligation, and you get an honest benchmark against your current planning.

Schedule a call

Edit content