Our new vehicle routing engine supports teams that need to automate and optimize the creation of vehicle routes as they scale. Our collection of pre-built constraints makes it easy to launch a vehicle routing app in minutes.
Start with a low-code SaaS solution for standard routing problems and then build on that exact same model to customize constraints and value functions, implement version control in your dev pipeline, perform experiments like acceptance tests, and more.
Launch the Nextmv Routing app, a low-code entry point
The Nextmv Routing app is a SaaS solution for route optimization powered by our new routing engine. Getting started is free and speedy:
- Create an account
- Select the Nextmv Routing app and start a trial
- Add the Nextmv Routing app
- Make your first API call to run the app
From there you can configure constraints like precedence and time windows. All it takes is a few clicks to get access to the route optimization API and start making calls.
A full suite of configurable vehicle routing constraints
Setting up and defining a vehicle routing model can be a barrier to entry for solving real-world use cases. We have been the developer or data scientist that finds, fiddles with, and implements algorithms specific to capacitated vehicle routing problems (CVRP) or capacitated vehicle routing problems with time windows (CVRPTW). We see forum posts with questions like “How do I mathematically model pickups and deliveries?” or “How do I bundle pairs of trips?” or “How do I add start and end locations to individual vehicles?”
We’ve removed this friction with a turnkey engine that exposes a suite of common constraints that are ready to use, off-the-shelf:
- Compatibility (identify which stops are compatible with which vehicles)
- Precedence (add pickups and deliveries or specify multiple pickups before deliveries and vice versa)
- Maximum route distance or duration
- Replanning / initial stops (add new stops to previously planned routes)
- Service time windows
- And more!
To use a constraint, simply enable it in the Nextmv Routing app by including it in your input file. See the example below for precedence:
A searchable, sortable history
Want to take a look at a specific route on a map to visually explore why some drivers have shorter routes than others? Need to review the input for a run that was made yesterday to evaluate that change you made in service time windows? Explore and troubleshoot in the Nextmv console.
The Nextmv Routing app includes a holistic view of your runs. Filter and sort your runs by date, name, and more. Then click into a run to see details including the output itself, the metadata associated with each run, the input used, and the route for each vehicle displayed on the map. Invite coworkers to your team so you can easily share links to a specific run or view within the console. Collaboration, for the win!
Go from SaaS to custom using the same model
While the Nextmv Routing app addresses many routing use cases, we make it possible to take your vehicle routing game up a notch. Using the same model that’s running behind the scenes powering our SaaS solution, you can create a completely custom model. Don’t change your data pipeline, don’t start from scratch.
Get all of the features mentioned above plus the ability to build custom constraints (e.g., package type to ensure that there’s a dedicated vehicle for packages from a specific vendor) and value functions (e.g., maximize fleet utilization) to pair with the out-of-the-box functionality already provided.
Create and deploy custom routing apps
SaaS routing solutions are easy to deploy, but are usually not customizable. Custom routing solutions are often hard to deploy, but very customizable. Nextmv gives you a path to travel between both worlds. With a few commands, you can publish a custom routing model you locally developed, and have your own API endpoint (it’s like your own custom SaaS!) that’s manageable from the Nextmv console.
Test your apps using our built-in experimentation platform
When you create and deploy a custom app, you get the console experience (with run history as described above) plus access to experimentation. This allows you to conduct and analyze experiments on your custom apps so you can feel confident about how app updates will affect your KPIs.
Answer questions that have operational impacts and incorporate model testing into your dev workflows. Perform batch experiments to explore the impacts of updates to your output, and then use acceptance tests to ensure that your candidate model is ready to go live in production.
Have questions? Want to chat about your use case? Reach out to us at any time.