NGC | Catalog
CatalogResourcesRoute Optimization Service Notebooks

Route Optimization Service Notebooks

Logo for Route Optimization Service Notebooks
Address the vehicle routing problem (VRP), optimizing for cost and time, with dynamic constraints.
Latest Version
October 10, 2023
Compressed Size
71.09 KB


The route optimization AI workflow demonstrates how to use NVIDIA cuOpt to minimize vehicle routing inefficiencies by finding the most optimal route for three use cases:

  1. Last mile delivery - fleets of trucks dispatched from distribution centers or depots to stores or end customers.
  2. Pickup and delivery - fleets of trucks dispatched to fulfill orders to multiple locations. For example, a DoorDash driver picking up a food order from a restaurant and delivering it to a customer’s home.
  3. Dispatch optimization - a service provider is dispatched to fulfill a list of service requests which are allocated a specific amount of time per request, which varies by request. For example, a telecommunications technician is dispatched to a customer’s home to install a router, and then to a different customer’s home to install a data cable.

The NVIDIA route optimization AI workflow includes details on how to deploy a sample AI solution for route optimization and includes origin-destination cost matrix creation, data preprocessing, the NVIDIA cuOpt Cloud Service GPU-accelerated solver pipeline, driving directions, and map visualization.

The NVIDIA route optimization workflow uses the cuOpt cloud service API to generate routes, enabling organizations to use cuOpt without needing to install, deploy, and manage it themselves. The cuOpt cloud service minimizes the development effort and simplifies the implementation, enabling organizations to leverage cuOpt – even without deep expertise.


Key benefits of the route optimization AI workflow include:

  • Dramatically reduce travel times and fuel costs with dynamic rerouting – which can save millions of dollars.
  • Route packages up to 120x faster, while maintaining the same level of accuracy, and reroute dynamically when unexpected issues arise.
  • Leverage GPU acceleration to scale resources and address the most demanding workflows, enabling use cases previously impossible to take on.

Getting Started

To get started, review the documentation linked below and learn what is included in the route optimization workflow, and how to run the workflow.

Have an upcoming accelerated optimization project? Apply to try the route optimization service AI workflow on NVIDIA LaunchPad.


Additional Resources

Learn more about how to use NVIDIA cuOpt through our Deep Learning Institute. Access the course here.

Contact NVIDIA to learn more about options for accessing the route optimization AI workflow and NVIDIA cuOpt.


By accessing the cuOpt Cloud Service, you accept the terms and conditions of this End User License Agreement.