Configuring Nextmv routing app

Configuration is defined using profiles in Nextmv Cloud. Profiles let you define a configuration once for settings that are consistent across runs, and use it within a run by referencing the profile identifier. For certain types of functionality you'll need to define a run profile in order to be able to use the feature. For example, some integration profiles require you to specify a secret API key that you wouldn't necessarily want to provide repeatedly. With a run profile, you can set it once so we can store and manage the key securely.

We support two types of profiles: run profiles and integration profiles.

• A run profile lets you configure information that impacts the behavior of a run.

• An integration profile lets you configure information used for integrations.

Run profiles let you configure information that impacts the behavior of a run but does not vary run to run, for example the type of measure used. A run profile optionally contains references to zero or more integration profiles. To create a run profile, send a request to the POST /v1/routing/profiles/run endpoint with a JSON payload that follows this schema:

Each integration object follows this schema:

A sample schema and response are provided below:

{
   "name":"Sample Run Profile",
   "id":"onfleet-run-profile",
   "description": "Delivery profile",
   "integrations":[
      {
         "type":"onfleet",
         "id":"my-onfleet-integration"
      }
   ]
}

Make sure to store the id, as you will need it when creating a new run. Please note that you may set up multiple run profiles using this method. You only need to do this process once for each API key.

To modify an existing run profile, you can use the PUT /v1/routing/profiles/run/{profile_id} endpoint with a JSON payload that follows the same schema as when creating a new run profile (described in the table above). Note that you may not modify the original id.

Onfleet is a logistics software focused on dispatch and fleet management. You can leverage the power of Nextmv's cloud API with the Onfleet integration.

Follow the steps below to configure the Onfleet integration.

1. Create the tasks and workers you want to route. To leverage the full feature set of the Nextmv Cloud API, please check out the following sections:
   • Compatible features. Onfleet's API functionality that is compatible with Nextmv Cloud API.
   • Extensible features. Use Nextmv Cloud API features not supported by the Onfleet API.
2. Get the Nextmv and Onfleet API keys. This only needs to be done once.
3. Set up an Onfleet integration on Nextmv to securely store the Onfleet API key.
4. Create a run profile for using the Onfleet integration when requesting new runs, using the integration id and type obtained in the previous step. This only needs to be done once.
5. Send a request to the /v1/integrations/onfleet/run endpoint with a JSON body that follows the input schema to route and assign the desired tasks and workers, using the run profile id obtained in the previous step with the run_profile key. The API server should return a run_id that you will use to query your results. This means that a Nextmv solver has been spun up to solve the model with the given input. The time it takes to finish execution varies by the size of the input and solver options.
6. Poll the status of your run with the /v0/run/{run_id}/status endpoint, using the run_id. Once the status is succeeded, you can query your results. Please see our section on using the API for more information.
7. Get the result of your run with the /v0/run/{run_id}/result endpoint, using the run_id. The result holds the Onfleet containers that represent the final assignments of tasks to workers and it will follow the structure of the output schema. Please see our section on using the API for more information.
8. Update the container's tasks in the Onfleet API using the results optimized by the Nextmv Cloud routing engine.
9. Manage your configurations using the Onfleet endpoints.

The Onfleet integration lets you securely store one or more Onfleet API keys inside of Nextmv Cloud. Send a request to the /v1/routing/profiles/integration/onfleet endpoint with a JSON payload that follows this schema:

The configuration object allows you to securely store your Onfleet API key and optionally set up the compatibility feature explained below.

The compatibility object allows you to fine-tune which task and worker attributes are automatically parsed when supported by compatible functionality. Using the compatibility feature is optional. Any attribute that is not explicitly defined as false will be automatically parsed as true. The object follows this schema:

In the following two tables you will find the schema definition for the tasks and workers object, respectively.

• tasks schema (you can read more about compatible Onfleet task attributes here):

  Key	Description	Required	Sample value	Type
  complete_before_after	Parse the Onfleet task's completeBefore and completeAfter into a Nextmv stop's hard_window.	No	false	boolean
  dependencies	Parse the Onfleet task's dependencies into Nextmv stop's succeeds.	No	false	boolean
  location	Parse the Onfleet task's destination.location into the Nextmv stop's position. NOTE: if you turn this toggle off (false) you should make sure that you are specifying position via task metadata, otherwise your stop will not have a location and will cause an error.	No	false	boolean
  quantity	Parse the Onfleet task's quantity into the Nextmv stop's quantity.	No	false	boolean
  service_time	Parse the Onfleet task's serviceTime into the Nextmv stop's stop_duration.	No	false	boolean

• workers schema (you can read more about compatible Onfleet worker attributes here):

  Key	Description	Required	Sample value	Type
  active_task	Parse the Onfleet worker's activeTask to be part of the Nextmv vehicle's backlog.	No	false	boolean
  capacity	Parse the Onfleet worker's capacity into the Nextmv vehicle's capacity.	No	false	boolean
  location	Parse the Onfleet worker's location into the Nextmv vehicle's start position.	No	false	boolean
  tasks	Parse the Onfleet worker's tasks to be part of the Nextmv vehicle's backlog.	No	false	boolean

A sample schema and response are provided below:

{
  "name": "My Onfleet Configuration",
  "description": "my onfleet configuration for deliveries in LA",
  "id": "my-onfleet-integration",
  "configuration": {
    "api_key": "b4**************************8a7",
    "compatibility": {
      "tasks": {
        "quantity": false
      },
      "workers": {
        "active_task": false,
        "tasks": false,
        "capacity": true
      }
    }
  }
}

Make sure to store the id and type, as you will need them to add the integration to your run profile. Please note that you may set up multiple integration profiles using this method. You only need to do this process once for each API key.

To modify an existing Onfleet integration, you can use the PUT/v1/routing/profiles/integration/onfleet/{profile_id} endpoint with a JSON payload that follows the same schema as when creating a new integration (described in the table above). Note that you may not modify the original id.

Onfleet's API offers functionality that is compatible with Nextmv Cloud API. If compatible attributes are defined for tasks and/or workers, they will be parsed to Nextmv features. In the following tables you will find a list of these attributes.

Task compatible attributes

Worker compatible attributes

To use Nextmv Cloud API features not supported by the Onfleet API, you may extend tasks and workers using Onfleet metadata. Metadata for tasks and workers is defined as an array of metadata objects. The name of each metadata object should follow this pattern:

"nextmv:<version>:<attribute>"

Where version dictates the metadata version used and attribute is the name of the Nextmv attribute used to extend the task or worker. An example of a name is: "nextmv:0.1:unassigned_penalty", used to set an unassigned penalty for a task (stop). You may use all (or none) supported metadata extensions for both tasks and workers. In the following tables you will find a list of these attributes.

Task extended attributes

Worker extended attributes

HERE is a platform that provides a number of location-based APIs. The Nextmv Cloud API can leverage HERE's matrix routing service to make routing decisions informed by street routes and real-time or projected traffic conditions. Nextmv's HERE integration allows you to construct measures using HERE routing data, providing more accurate routes than the haversine measure.

Follow the steps below to use the HERE integration.

1. Get the Nextmv and HERE API keys. This only needs to be done once. Note, The HERE API key will need to be associated with a plan that has access to the HERE matrix API.
2. Set up a HERE integration on Nextmv to securely store the HERE API key. We will securely store your API key and use it when you request new runs. This only needs to be done once.
3. Create a run profile for using the HERE integration when requesting new runs, using the integration id and type obtained in the previous step. This only needs to be done once.
4. Follow the Nextmv Cloud API reference guide to create a new run, using the run profile id obtained in the previous step with the run_profile key. When using a run profile with the HERE integration, you may also specify additional information to incorporate traffic. After, you can poll the status of your run and get the result.
5. Manage your configurations using the endpoints.

You can specify real-time or historical traffic data when using the HERE integration. Traffic data is specified in the run_profile field of your JSON input when making a new run using the departure_time field (see example below). Specifying a departure_time for your vehicles will cause HERE to factor traffic conditions into its route lengths in one of two ways:

• If you set departure_time to current time, HERE will use real-time conditions.
• If you set departure_time to a future time, HERE will use estimates based on historical data.

The format for the departure time must follow an RFC3339 formatted timestamp.

{
  "run_profile": {
    "id": "my-here-integration",
    "measure": {
      "here": {
        "departure_time": "2018-12-19T18:23:45"
      }
    }
  }
}

See the HERE documentation for more information on historical and real-time traffic information. Note that we do not support the any parameter used by HERE. If you want your route lengths to be calculated without considering traffic, simply omit the departure time parameter from the run profile. If you want to use current traffic conditions, provide a timestamp for the current time.

Your HERE account will incur charges for each run of a Nextmv model that uses a run profile configured with the HERE integration. Make sure you understand the cost implications of the HERE API before you start using the HERE integration. We recommend testing against the Nextmv API using another measure, such as the Streets or haversine (default) measures to avoid incurring significant costs during development.

For each run, Nextmv will request a matrix of routes where the number of sources and destinations are equal to the number of stops plus the number of vehicle stop and end locations. For example, if your run has 1000 stops and 20 vehicles, each with start and end locations, we will request a 1040x1040 matrix.

HERE charges based on the number of "transactions" your account performs, as described in the HERE documentation. The exact cost per transaction depends on your HERE account's plan.

The HERE integration lets you securely store one or more HERE API keys inside of Nextmv Cloud. Send a request to the /v1/routing/profiles/integration/here endpoint with a JSON payload that follows this schema:

A sample schema and response are provided below:

{
  "name": "My HERE Configuration",
  "id": "my-here-integration",
  "configuration": {
    "api_key": "b4**************************8a7"
  }
}

Make sure to store the id and type, as you will need them to add the integration to your run profile. Please note that you may set up multiple integration profiles using this method. You only need to do this process once for each API key.

To modify an existing HERE integration, you can use the PUT/v1/routing/profiles/integration/here/{profile_id} endpoint with a JSON payload that follows the same schema as when creating a new integration (described in the table above). Note that you may not modify the original id.

Streets is an integration for extending Nextmv measures using real road networks when calculating the distance between two points. While the haversine measure can quickly estimate the cost of traveling between two points, this measure can be limited in its accuracy as it is based on "as the crow flies" distance. The Streets measure takes into account conditions such as one-way roads, bodies of water and speed limits when calculating distances and durations. Vehicle profiles are used to determine speed, so vehicle speeds do not need to be specified in the input file.

Follow the steps below to use the Streets integration.

1. Contact support to request access to the Streets integration alpha.
2. Once your account has been granted access, get your Nextmv API key and set up a Streets integration. This only needs to be done once.
3. Create a run profile for using the Streets integration when requesting new runs, using the integration id and type obtained in the previous step. This only needs to be done once.
4. Follow the Nextmv Cloud API reference guide to create a new run, using the run profile id obtained in the previous step with the run_profile key. After, you can poll the status of your run and get the result.
5. Manage your configurations using the streets endpoints.

Send a request to the /v1/routing/profiles/integration/streets endpoint with a JSON payload that follows this schema:

A sample schema and response are provided below:

{
  "name": "My Streets Configuration",
  "id": "my-streets-integration"
}

Make sure to store the id and type, as you will need them to add the integration to your run profile.

To modify an existing Streets integration, you can use the PUT/v1/routing/profiles/integration/streets/{profile_id} endpoint with a JSON payload that follows the same schema as when creating a new integration (described in the table above). Note that you may not modify the original id.