As promised earlier this year, we’re very happy to finally announce the unveiling of our new route optimization endpoint!
We deployed an instance of the popular Vroom open-source engine, which is capable of solving complex Vehicle Routing Problems (VRP) in record time. This type of problem always occurs when multiple locations need to be visited in the optimal order by one or more vehicles. Consequently, it’s most valuable for logistics planning, but is also useful for traveling sales persons (which actually is the name of a particular VRP). With Vroom job and vehicle scheduling is a breeze.
The optimization service supports advanced parameters to constrain the optimization, such as:
- capacities: each
vehiclecan have separate capacities for multiple goods, eachjobwill consume a vehicle’s capacity - time windows: each
vehiclecan have a start and end time (e.g. working hours), eachjobcan have multiple time windows, expressed as week seconds, e.g. Mon 8 am = 28800 - skills: each
jobcan require skills thevehiclemust meet - service duration: each job can take a specified amount of time
Optimize two vehicles for 6 jobs in Berlin
The full documentation how to use this endpoint can be found in our API documentation or on Vroom’s Github page.
The optimization works with all available profiles (car, various bike variants, pedestrian, wheelchair and more) of OpenRouteService. So far, it has only been implemented in the Python SDK, but will soon be available in the JavaScript, R, and QGIS clients as well.
See also our related research on healthy, quiet and green routing, wheelchair accessibility, Landmark navigation or routing through open spaces and more.
Happy optimizing!