Vehicle Routing Problem (VRP): Definition, Variants, and Real-World Examples

What is the Vehicle Routing Problem (VRP)?#

VRP is the general “fleet routing” problem behind delivery routing, field service scheduling, and multi-driver route planning. It asks: how do you build one or more routes that visit all required stops efficiently?

In practical terms, VRP means deciding:

• Which vehicle should serve each stop
• In what order each vehicle should visit its assigned stops
• When each stop should be visited (ETAs), while staying within constraints

This is exactly what route optimization software does. If you want the practical overview, start with the pillar: Route Optimization: Complete Guide.

VRP vs TSP: what’s the difference?#

A common confusion is VRP vs the Traveling Salesman Problem (TSP).

• TSP: one vehicle must visit all stops once and return (or end) — “best single tour.”
• VRP: multiple vehicles may exist, stops must be assigned to vehicles, and constraints are common (time windows, capacity, working hours).

In other words: VRP is “TSP for the real world.”

Why VRP is difficult#

VRP becomes hard quickly because the number of possible solutions explodes: you’re not only choosing an order of stops, you’re also choosing which vehicle serves which stops.

Add constraints like time windows and capacity and many candidate routes become invalid. That’s why production route optimizers focus on high-quality solutions quickly, not perfect optimality.

Common VRP variants (types you’ll see in practice)#

Most “route optimization” problems are one of these VRP variants (or a combination).

If your business operates on appointment windows, see: Route Optimization with Time Windows. If you use multiple drivers, see: Route Optimization with Multiple Vehicles.

Real-world constraints VRP models support#

In real operations, the “shortest route” is often not feasible. Constraints define what’s allowed.

• Time windows: arrive only during customer availability
• Service time: time spent at each stop (delivery, inspection, paperwork)
• Vehicle capacity: weight/volume limits and stop demand
• Working hours: shift start/end for each vehicle/driver
• Start/end locations: depot, warehouse, or driver home base
• Priorities: urgent stops first or guaranteed delivery rules

This is why “stop reordering” is not the same as optimization for fleets. For the practical workflow, read: How route optimization works.

How VRP is solved in practice (without heavy math)#

Large VRP instances are rarely solved by brute force. Instead, route optimization engines typically:

• Build an initial solution quickly (a feasible route plan)
• Improve it iteratively by swapping stops, moving stops between routes, and re-ordering sequences
• Balance speed vs quality to return results fast enough for daily dispatching

Practical takeaway: the best routing tools focus on feasible routes that drivers can execute — not just “shortest distance.”

If you want the business outcome side, see: Route optimization benefits.

What data you need to solve VRP#

To get accurate routes and ETAs, you need good inputs. Here’s the minimum checklist:

Stops

• Address or latitude/longitude
• Optional: service time (minutes at stop)
• Optional: time windows (earliest/latest)
• Optional: demand (weight/volume)

Vehicles / drivers

• Start location (and optional end location)
• Working hours (Time In / Time Out)
• Optional: capacity limits

TrackRoad supports these inputs (including Excel import) here: Route Optimizer.

Simple VRP example#

Imagine 30 delivery stops and 3 vehicles. The optimizer must decide which stops go to which vehicle, then order the stops on each route to minimize travel time, while ensuring each driver finishes within working hours and time windows are met.

That combined “assignment + ordering + constraints” is VRP — and it’s why route optimization is more than navigation.