Mapbox software solutions offers SDKs and APIs that facilitate the operation of a micromobility platform. 

In an era where urban mobility is rapidly evolving, micromobility solutions such as e-scooters, e-bikes, and shared bicycles are becoming indispensable for efficient and sustainable transportation. At Mapbox, we are proud to support this transformation through investing in our innovative SDKs and APIs. These tools enable micromobility platforms to offer precise, reliable, and user-friendly services, enhancing the urban mobility experience for millions of users worldwide. In this blog, we’re going to cover how Mapbox empowers micromobility platforms by providing technical details on our innovative SDKs and APIs.

Mapbox for Micromobility

Micromobility platforms rely on accurate and real-time data to operate efficiently and safely. Mapbox provides the essential mapping and location technologies that underpin these services. Here's how Mapbox empower micromobility platforms:

Dynamic and Real-Time Mapping

From our APIs for route calculation, isochrone creation, ETA determination, search over multi-modal routing capabilities (walking, driving/cycling, walking) to Maps SDKs for both mobile and web applications and Navigation SDKs for turn-by-turn navigation, Mapbox provides many solutions for micromobility companies.

Enhanced User Experience

Together with our Optimization API, these can be used to enable users to search for vehicles, destinations and parking locations, guide them and bill them, but also to plan optimal work distribution for maintenance jobs, and to guide and track maintenance orders via turn-by-turn navigation. 

Fleet Operations

Our web SDK Mapbox GL JS enables developers to build dashboards that gives administrators insight into the current state of operations and could allow them to make critical business decisions in real time. Mapbox has performant hosting capabilities for map data such as parking locations, no-drive zones, custom POIs and many more.

Additionally, Mapbox is currently working to bring more great features for micromobility companies including server side routing engine implementations for polygon exclusion, electric bicycle routing, weather implementation, and voice interaction capabilities.

Micromobility Architecture Overview

To better understand how Mapbox integrates with micromobility platforms, let’s take a look at the following architectural drawing. To understand which Mapbox products make up the components of the architecture, please refer to the detailed architecture description below the diagram.

Architecture Components

Embedded Application

The embedded application runs on the vehicle*, has access to the vehicle's positioning device, and communicates with the backend to update it for asset tracking purposes. It can lock and unlock the vehicle and report its status. For example, X uses X for their embedded application.

*Note: This is not something Mapbox provides. It is an app running on the OS of the vehicle itself and communicates where the vehicle is and what its status is to a backend.

Proposed Mapbox components for this application: None

User App (iOS and Android)

The user application is the user-facing client application. It allows users to sign up for the micromobility services, use the services, get support, and pay. 

Proposed Mapbox components for this application:

• User authentication
• Find close vehicles:
  • Backend: Mapbox Isochrones API to create polygon to compare all vehicle locations with (show all vehicles within a 10 minute walk) or
  • Backend: Mapbox Matrix API to determine the distances and ETAs to all vehicles in a defined boundary (show all vehicles in my postal code)
  • Mapbox Maps SDK (iOS / Android) to display vehicles and user location on a map
• Multi-modal route planning:
  • Mapbox Search API for inputting destination
  • Multi-modal navigation, respecting No-Drive-Zones and No-Parking-Zonessome text
    • Mapbox Directions API, brute force approach for Multi-Modal
    • Mapbox Navigation SDK (iOS / Android)) for turn-by-turn (TbT) navigation on proposed multi modal route, using different movement profiles
• Vehicle linking to user via communication with backend DB
• Direct link to support team in case of problem / emergency situation
• Billing (Printing bill):
  • Mapbox Static Images API: The traveled trip on a map as part of the bill
  • Mapbox Navigation SDK enhanced location engine (iOS / Android) to mitigate incorrect billing via GPS inaccuracies that would result in additional (incorrect) travel distance
• Optional: Voice interaction capability

Admin Dashboard for Monitoring and Reporting (Web Application)

The admin dashboard is a web application that enables decision makers to make critical decisions in real time, based on the current position, orientation, and speed of assets. It also allows monitoring the maintenance teams work progress and for manual intervention in case there is a business need.

Proposed Mapbox components for this application:

• Mapbox GL JS: Overview of different regions, monitor vehicles, maintenance team work, parking situation, distribution of vehicles

Backend service

The backend holds database and services for billing calculation and fleet operations and rebalancing.

Proposed Mapbox components for this application:

• Database
  • Vehicle live positions
  • Parking availability
  • Maintenance team asset locations
  • Maintenance jobs
• Serve vehicle locations on requests of user app, optimization service, maintenance app 
  • Mapbox Isochrones API to compare vehicle locations with isochrones polygon before responding to requests
• Serve maintenance jobs to maintenance app
• Billing calculation:
  • Mapbox Directions API in case of active route and distance based billing
  • Mapbox Map Matching API in case of no active route and distance based billing
• Issue of the bill:
  • Mapbox Static Images API for the traveled trip on a map as part of the bill
• User authentication

Fleet Operations and Rebalancing 

Fleet operations and rebalancing is part of the backend and automatically plans maintenance and re-distribution jobs for the fleet operations team (maintenance). This service can run in real-time and update the schedules based on live traffic conditions, vehicle distribution, and maintenance needs.

Proposed Mapbox components for this application:

• Maintenance trip optimization:
  • Mapbox Optimized Trips v2 to solve the constrained vehicle routing problem to optimize and plan maintenance and re-balancing jobs
  • Mapbox Matrix API to solve the underlying cost matrices that take live traffic into account, in case of a custom optimization algorithm)
  • Mapbox Directions API to returns the actual route on the road network to follow after optimizing the waypoints of a maintenance job

Maintenance Application

The maintenance application is an internal application for members of the fleet operations team who re-distribute and maintain the vehicles, so the user experience stays great.

Proposed Mapbox components for this application:

• User authentication
• Retrieve jobs for authenticated user
• Display of jobs:
  • Mapbox Maps SDK (iOS / Android)
  • Mapbox Directions API
• Navigation on maintenance jobs:
  • Navigation SDK (iOS / Android)) for TbT navigation on maintenance job 
• Asset tracking:
  • Reporting on status of jobs
  • (Optional) additional asset tracking for maintenance vehicles via separate app

Map Data Hosting

While the backend hosts a database with live vehicle locations and maintenance asset locations, the map data hosting is meant for less frequently updating data that is served to end users. We differentiate such data in fast updating data and slowly updating data.

Proposed Mapbox components for this application:

The Mapbox TIling Service (MTS) can be used for data that updates less frequently than every 12 seconds. Through Mapbox data hosting and tiling, the data will be served only to those users who are displaying it on their map. Also the highly scalable Mapbox infrastructure, which supports >700 million monthly active users, guarantees high performance and a great user experience.

• Example for fast updating data:
  • Parking availability

• Examples for slowly updating data:
  • No drive zones
  • No park zones
  • Slow drive zones

Multi-modal routing

As part of the Mapbox product offering, multi-modal routing is targeted at micromobility providers, among others. The goal is to provide a seamless navigation experience by integrating different transportation modes into a single journey. 

Examples of micro mobility multi-modal routing:

• Walking to scooter/bicycle pickup: The multi-modal routing service begins by calculating a pedestrian-friendly path from the user’s current origin point to the nearest location where they can pick up an electric scooter. This would take into account sidewalks, pedestrian zones, and safe crossing points to get the user to the scooter safely and efficiently.
• Scooter/Bicycle leg: Once the user has reached the scooter or bicycle the next leg of the journey involves routing from the vehicle pickup location to a designated vehicle parking area near the user's final destination. This part of the journey would be optimized based on factors pertinent to scooter travel – for example, bike lanes, scooter-friendly routes, and potentially restricted areas where micro mobility vehicles might not be permitted. The goal is to provide an efficient and legal route that takes advantage of the scooter’s ability to navigate different types of roads and paths that may not be accessible to cars.
• Walking from scooter parking to destination: After parking the scooter, the user would need directions to walk from the parking location to their ultimate destination. Again, like the first leg, this path would be optimized for pedestrians, considering aspects like the shortest safe routing, walking time, and ease of navigation.

Mapbox's multi-modal routing takes into account real-time traffic conditions, local regulations, and infrastructure to provide an optimized route. It uses extensive bespoke geospatial data and algorithms to make different transportation modalities work together in a cohesive routing experience. 

For micromobility providers, integrating multimodal routing can enhance their service offering by providing their users with a comprehensive, end-to-end navigation solution within a single interface.This functionality becomes increasingly important in urban areas where traditional vehicular traffic can be congested, and where users might prefer faster or more environmentally friendly modes of transportation for parts of their journey. By providing accurate, up-to-date routing that combines walking and 2-wheel conveyance, Mapbox helps micro mobility providers offer more efficient and attractive services to their customers.

Technical implementation

The implementation of multi modal routing is a client side solution. To facilitate multi modal routing, we use the Mapbox Directions API and Matrix API. The client side solution is a three-step process:

1. Find locations where the mode of transport can be changed, both around origin and around destination.
2. Get the cost matrix for walking mode from origin to the first set of mode change locations, the cost matrix between the first set of mode change locations and the set of mode change locations around the destination, and the cost matrix between last set of mode change locations and the destination location.
3. Compute all possible permutations and choose the combination of the 3 legs with the lowest overall cost.

This process is a brute force approach which yields the optimal solution, given the input. The matrices required for this approach are:

• 1 × sizeOfOriginSet
•  sizeOfOriginSet × sizeOfDestinationSet
•  sizeOfDestinationSet × 1

To use the optimal combination, the Mapbox Directions API is applied to generate a routable ‘route object’ that can be fed to the Mapbox Navigation SDK and can be displayed on a Mapbox map.

To do so, Directions API requests are made for each of the identified legs of the calculated combination. The first request is made using the movement profile of the first legs mode of transport, the second and the third with their respective modes of transport.

Example of implementation

The code link below is a client side implementation of the multi modal routing concept. It fetches the mode of transport change locations for a tileset using Mapbox Tilequery API. These requests can be exchanged by requests to other APIs that serve the mode of transport change locations.

See code here: https://codepen.io/Moritz-F/pen/RwmovZO 

Note: To make the below code work, please exchange <YOUR_ACCESS_TOKEN> by your Mapbox access token, and add a tileset URL <YOUR_TILESET> that holds the parking and pickup locations.

Ready to revolutionize micromobility? Contact our Sales team today to unlock the power of Mapbox for your platform.