TODAY, IT IS NO LONGER a question of whether cars will drive themselves, but rather when this will happen at scale.

Ford Motor Co. announced last month that driverless cars are coming to the streets of our nation's capital. With this latest addition, Ford will be piloting autonomous vehicles in Washington, Detroit, Miami and Pittsburgh. Add just one other company – Waymo – and the map of corporate pilot programs expands to 15 Californian cities, six Arizona cities and to Austin and Atlanta and beyond. By one count, at least 46 companies are piloting these vehicles in cities across the country and the world. Even cities themselves are getting into the game by testing autonomous shuttles. The way we all get around is transforming as firms invest billions of dollars to develop, test and deploy autonomous vehicles on urban streets.

Fifty percent of large American cities are now exploring how to integrate self-driving vehicles into their long-range transportation plans – up from under 10 percent in 2015. Before long, autonomous vehicles will be ubiquitous on our roadways, but the full story has not yet been written.

Cities need to be in the driver's seat as we transform from the current mobility environment to our autonomous future. In order to help city leaders do this, our latest National League of Cities report, Autonomous Vehicle Pilots Across America, provides guidance and recommendations on successful autonomous vehicle program approaches.

This research shows that cities must first determine their goals and reasons for pursuing a self-driving pilot project to make sure it is scoped appropriately for their community. Engaging the private sector and creating a consortium of local partners within the private, nonprofit and regional government sectors are key steps. Then, deciding how the project should be phased in – all at once or in stages – becomes central, as well as working together with the state, which holds a great deal of sway in how plans are implemented and carried out. These steps will help ensure that autonomous vehicles support — rather than exacerbate — existing city goals around mobility, sustainability and land use. The people side always needs to take precedence over technology.

We see examples of cities leading on this front all across the country. Three unique cases are in Pittsburgh; Chandler, Arizona; and Arlington, Texas. In each city different models are underway, with leadership, experimentation and partnership serving as critical differentiating factors.

In Pittsburgh, the relationship between the mayor's office, Carnegie Mellon University and private sector partners spurred the development of a strong, well-aligned coalition for autonomous vehicle leadership.

Carnegie Mellon is a talent engine for Pittsburgh – and one of the key centers of the global autonomous vehicle ecosystem – attracting and spinning off thriving startups. Early on, Uber partnered with Carnegie Mellon's robotics department to launch a fleet of self-driving cars on city streets. This relationship soured as Pittsburgh leaders felt the company wasn't meeting its side of the bargain, backing away from initial promises on free rides and job creation, among others. But it's not the only self-driving car company in town. Ford-backed Argo AI is testing self-driving vehicles in Pittsburgh, as is autonomous car startup Aurora. And Aptiv – Lyft's autonomous partner – may be moving into a 70,000 square foot space in the city. These companies are all global leaders in the autonomous vehicle ecosystem.

Propelled by a wave of enthusiasm and support from Arizona Governor Doug Ducey, Chandler welcomed a city-wide pilot with Waymo in 2016. A robust tech ecosystem with companies like Intel, Garmi and several auto manufacturers made Chandler a natural choice for Waymo – but it also laid the foundation for a welcoming public. Several thousand residents showed up to Waymo's initial demonstration in 2015, kicking off a reciprocal and participatory relationship between the company and the public.

Like Pittsburgh, the city has yet to reach formal agreements with autonomous vehicle developers, focusing instead on facilitating collaborations with public departments. Though autonomous vehicle regulation is primarily happening at the state level, Chandler is pioneering ways to prepare for an autonomous future. The city council amended the zoning development code to reduce parking space by up to 40 percent when demand falls and encourage passenger loading zones as drop off and pick up locations for autonomous vehicles.

Arlington took the bull by the horns with its approach to self-driving vehicles. In an experiment to understand the technology, familiarize the public and signal Arlington as a testing ground for innovative research and deployment, the city council leased two autonomous shuttles to carry passengers between parking lots and sporting and concert venues.

The shuttles are wheelchair accessible with an onboard operator, and the fixed routes are closed to other traffic. The success of this pilot fed directly into plans for a second phase of testing with mixed traffic. Though Arlington's model shifts costs to the city, it also increases control over implementation – something other cities cannot do in the same way with fully private sector testing.

As we observe these pilots on the ground, questions about the next iteration of autonomous vehicles become inevitable. What will the soon-to-be-felt real-world changes look like as we move from pilots to full-scale implementation?

In order to explore this future, we developed autonomous vehicle scenarios with our partners at Bloomberg Philanthropies and the Aspen Institute that cities can use to plan now. This series of autonomous vehicle scenarios is built for city leaders and community members to explore long-lasting impacts on mobility, sustainability, jobs, the economy and urban transformation.

One particular example is a scenario that centers on how self-driving technology will broadly affect urban transportation methods – as foreshadowed by the Arlington shuttle experiment – like buses and subway systems. Scaling from single vehicle pilots, cities along with the private sector could potentially build networks of shared autonomous minibuses.

Shuttles could serve equity goals across the spectrum, from helping lower-income community members to disabled city residents and others who live in "transit deserts." These services may also offer a solution for those who don't have a car to travel to opportunities that are farther away.

With broad deployment of autonomous vehicles on the near horizon, cities have a unique opportunity to reshape urban transportation and make it more people-centered, flexible and responsive. The best way to understand how these self-driving cars will look in our cities is to test them on the streets.