When I talk to a mayor’s office about autonomous vehicles, more often than not they want to know how they can get vehicles on their city’s streets as soon as possible. The rationale generally centers on one of two themes: (1) “the halo effect” – the perception that AV testing will lure innovation jobs and dollars to the city; or (2) “the Instagram effect” – the need to keep pace with peer-cities whose mayors’ already have pictures riding in or cutting the ribbon on an AV pilot. What I hear cited much less frequently is “the utility effect” – a rationale tied to transportation or mobility needs: my city wants to use AVs to address ______ challenge. Perhaps not surprisingly, it’s been cities with this focus who have had the most success with AVs to date.

I live in Washington, D.C. While the road environment for bicyclists here is improving, I wouldn’t call it a biking city. However, year after year in the spring I see friends decide this will be the year they start biking to work. They go to the bike shop with a price range and come home having blown through that budget with an aluminum or carbon frame road bike, racing seat, deep section wheels, and clipless pedals. These friends’ themes for purchase will be familiar from the paragraph above (1) “the halo effect” – a nice bike that will mean getting in shape and/or losing weight or (2) “the Instagram effect” – their friends post pictures with cool bikes and they want to do the same (#BAAW, #WYMTM).

As with cities, too infrequently do I hear “the utility effect” cited: I want the option of biking for my commute/errands and bought a bike that fits that use case.

What do these cities (and to a lesser extent bicyclists) who’ve focused on the utility effect have in common? They approached new technology as a tool to meet their existing needs and set up key data-driven metrics for planning deployment and measuring success. This starts with the questions of “why do I want an AV?” and “how do I plan this technology?” and ends with a plan for how to measure if the deployment was successful. Last year I released a paper at the South by Southwest conference in Austin, Texas that introduced the concept of data-driven deployment for AVs, and INRIX has been working to evangelize this model with stakeholders since. At its core the report provided a model to use INRIX data about how and where constituents move through a city to match AV technology with system needs.

Is your city looking to meet a first mile/last mile need? INRIX data can identify where the highest concentration of those trips take place. Looking to address congestion in a city center? Our data can target routes with high concentrations of short-distance trips. Does your city want to provide new mobility access for seniors or disabled populations? Big data can help focus deployment on the routes these communities travel most. With any of these use cases, the key is establishing clear metrics at the outset to measure the success of the project and be prepared to refine if needed.

It shouldn’t be surprising the cities worldwide that have had the most success with AVs are those who have a clear plan to use the technology to address their specific transportation challenges.

This includes a city like Christchurch, New Zealand, where a 2011 earthquake (among many other damaged structures) rendered the hospital’s parking garage unusable. The city is now considering an HAV shuttle to ferry residents from an off-site parking lot to the hospital. It also includes a city like Boston, Massachusetts, where the city has laid out a clear timeline and milestones for expanded AV testing (by multiple operators) in the Seaport and committed to working with the World Economic Forum and Boston Consulting Group to understand and model what impact the technology could have on Bostonians’ travel. It’s these sorts of cities focused on AVs solving a particular problem who are finding the most traction in their testing and should serve as a model for others considering AVs on their own streets.

As with a friend considering the purchase of a new bicycle, I’d urge cities to consider what they want the technology to be used for and approach the selection process with a clear plan of how to invest given these goals. This means data-driven metrics and key performance indicators. Come fall in Washington, D.C. there are a lot of expensive, lightly-used road bikes for sale, while those friends who bought a less flashy around town bike (or joined a bikeshare program) are usually still riding regularly. It’s up to cities to support AV plans that have staying power and not new technology for the sake of new technology.