Case study: Axon Vibe

The Opportunity

Intermodal Routing Under COVID-19 Lockdowns

In summer of 2020, Axon Vibe came to Mobi with a COVID-related planning problem. The transit operator in a major U.S. city planned to close its subway system from 1 am to 5 am nightly for cleaning and disinfection. Essential workers who would normally use the subway during those hours needed alternative ways to travel between home and work.

In response, the transit operator launched a program that made essential workers eligible for one free trip in a for-hire-vehicle per night if a) a bus trip crossed a specific duration threshold, b) required more than two transfers, or c) required a walk of more than a half mile to or from a bus stop.

Axon Vibe was tapped to build a mobile app for this essential worker program. It needed a partner able to provide intermodal routing for a transit system that in 2019 served millions of riders with thousands of subway cars and buses. Importantly, this routing system would need to work around nightly subway closures, incorporate multiple modes of transportation, and honor the essential worker benefits. With expertise in operations research, fleet planning, and adaptive network optimization, Mobi was the right partner for the challenge.

The Solution

Customize to the city, personalize to the rider

Worldwide, public transit is often underutilized because conventional routing technologies struggle to meet the unique needs of transit operators and their riders. For example, some people are willing to walk or bike long distances; others may need wheelchair-accessible options. Some people have a tight budget, while others are willing to pay up for a faster journey. The essential worker program became a test case for delivering an intermodal routing solution that can adapt to diverse users and needs.

After studying the transit operator’s ecosystem alongside Axon Vibe, Mobi customized its Agent product, which contains an intermodal routing engine, to integrate with Axon Vibe’s transit apps.

Users enter a destination into their Axon Vibe app, then Mobi Agent finds the most efficient routes by referencing:

• Schedules for subways, trains, and buses.
• Coordinates for “switching points” such as subway and bus stations.
• Available modes of transportation between switching points.
• The user’s declared preferences and historical routes.
• Service interruptions and delays.

Within a second, Mobi Agent presents the best routes and welcomes the user to select one. Choice is important because people’s needs can change day to day. Commuting alone, for instance, is different from commuting with a load of groceries or young children.

Mobi Agent continuously monitors the user’s route from A to B. If a new delay or interruption would prevent the user from completing their route or significantly lengthen the journey, Agent generates alternatives in real time. Spared from the stress of rushing or replanning, users can trust that they are always on the most optimal route for their needs.

All Mobi solutions run on AWS. Mobi Agent uses Amazon Elastic Compute Cloud (Amazon EC2) instances to dynamically scale up or down depending on user demand. Further, each core Agent microservice runs in containers provisioned by Amazon Elastic Kubernetes Services (Amazon EKS), while all underlying Agent data is stored in Amazon Relational Database Services (Amazon RDS) and Amazon Simple Cloud Service (Amazon S3). Lastly, Axon Vibe consumes Mobi Agent solutions through Amazon API Gateway.