Recent headlines suggest that the self-driving cars of the future have arrived. Google sibling Waymo has rolled out paid autonomous taxi rides in Phoenix, San Francisco and Los Angeles, with more cities to come. Tesla unveiled their Robotaxi in October 2024 and plans to commence production in 2027.

This progress is enough to make city planners hopeful for an imminent transformation. But widespread adoption of autonomous vehicles (AVs) could be a long time coming. 

Before self-driving cars can live up to their potential, we need major changes to our physical infrastructure, laws and the way we think.

Currently, the U.S. sees about 40,000 traffic fatalities a year, mostly due to human error. Transportation emits more greenhouse gases than any other sector, and traffic jams cost drivers countless hours of lost time. 

AVs paired with well-designed roads can help solve these problems. But to achieve full effectiveness, cities need to incorporate sensor-enabled “smart roads”—dedicated AV lanes that synchronize vehicle movements and detect potential hazards. 

Getting those roads built is just one challenge among many.

For AVs to be widely accepted, they need to protect passengers better than human-driven cars. That requires the vehicle’s software to be trained on every possible situation it could encounter.

One approach is extensive real-world testing—logging billions of miles with sensor-equipped vehicles to map streets and refine object detection. But this method is slow. One estimate suggests AVs would need to drive 8.8 billion miles to gather the necessary data.

A more efficient alternative is visual simulation. By creating “digital twins” of physical roadways, companies can simulate real-world environments and test AVs in hazardous conditions without endangering real people.

Producing AVs at scale is another obstacle. Since they use new and complex technological systems, they require a more sophisticated production process. Virtual models and simulations help ensure these technologies work properly before building expensive prototypes.

A lack of clear regulations could also further delay AV adoption. In the absence of federal rules, companies will be left with a patchwork of conflicting state laws, making it difficult for manufacturers to navigate compliance.

Perhaps the greatest challenge will be agreeing on a shared set of ethics. While AVs won’t drive distracted or fatigued, accidents will still happen. Who bears responsibility in a crash—the car manufacturer, the software developer or the passenger? Even more difficult is deciding how AVs should behave when a collision becomes unavoidable. Should a car prioritize protecting its passengers or minimizing casualties? Would consumers buy a car that might sacrifice its own passengers?

Self-driving cars have the potential to create safer streets, increase mobility and make transportation more efficient. But to realize these benefits, we must confront these challenges with smarter infrastructure, clearer regulations and a shared ethical framework.