If you’ve been to a big city in Asia, Europe, or North America over the past few years, you may have seen them: the whirring, sensor-studded electric autonomous vehicles (AVs) that navigate city traffic on their own, signaling, merging, braking—all without a driver. Once a futuristic novelty, AVs are now quietly becoming part of daily life in urban areas worldwide.
Waymo, Alphabet’s self-driving-car unit, is a leader in the AV industry. The company now completes more than 250,000 paid rides a week across markets, including Austin, Los Angeles, San Francisco, Phoenix, and recent addition Atlanta; Dallas, Miami, and Washington, DC, are on deck, and international testing is underway in Tokyo.
At a recent McKinsey conference in California’s Carmel Valley, Waymo’s chief product officer Saswat Panigrahi, and McKinsey Partner Emily Shao discussed how AVs have transformed from a niche experiment to a mainstream transportation option. Their conversation has been edited for length and clarity.
Reaching beyond early adopters
Emily Shao: Let me start with an anecdote: I took my parents, who are in their 70s, on a Waymo ride last year in San Francisco. They are not your typical tech first adopters, but they loved it. The first question they asked me when they got out of the vehicle was, “When can we buy one of these?” How have you seen ridership diversify in recent years?
Saswat Panigrahi: I believe we are in a unique moment where the inflection point has arrived. It has been amazing to see all the love from diverse users. My in-laws, who are in their 80s, have also taken a ride.
We see that trust builds quickly based on the smoothness of the drive and how consistent the experience is. When you think about how many folks over 70 depend on others to get around, this is a huge enabler of their personal autonomy. That’s also true for visually impaired people.
AVs are great for parents just taking their kids around. When parents pick the kids up from soccer practice, they can look in their kids’ eyes and ask them, “How was your game?” as opposed to being focused on the road, stressed about driving, and only half-listening. Giving people this time back, this freedom—making our autonomous vehicles not just a privilege for a few but generally safe and accessible for all—is really enriching to watch.
I’ve personally experienced the benefits of all the time you get back when using an AV. When I’m in San Francisco and I have to shuttle between meetings in different parts of the city, in the past, that would have involved some parking-related stress, at the very least. Now I don’t have that. And I have my own space for sensitive work calls. I can use the Waymo as a mobile office. If it’s a late night, I can doze off in the car if I need a nap.
Or think about date night, riding in a taxi or rideshare. Sometimes, you can be in the awkward position of, “The date can begin when we get to the restaurant, but before that, we have to watch our conversation.” Now, you have not only privacy but also the freedom to have a couple of drinks and not worry about getting home.
Emily Shao: If someone says, “I’m nervous about getting into a self-driving car,” how would you win them over?
Saswat Panigrahi: In 2019, I spoke to the first ten riders to try a fully autonomous ride in Phoenix. What was surprising was that it didn’t take a lot of convincing. Folks were willing to give it a try. Then we interviewed them when they got out and saw how sold they were after just one ride. That was powerful.
In 2025, people have heard about Waymo’s safety record, or they’ll be influenced by a friend, colleague, or acquaintance who has tried riding in a Waymo. I’ve found that once they enter the car, everyone becomes a convert. Within a few minutes, they notice the car’s capabilities. Of course, we don’t rest on that. We continue to test ourselves at the scale of billions of miles of simulation.
Emily Shao: How will the driverless experience change the design of the cabin interior? Is there a path toward a custom-built vehicle where there is, for example, no steering wheel?
Saswat Panigrahi: There’s a lot of room for innovation in the interior of the cabin now that it’s being designed for a rider instead of a driver. For decades, cars have largely been designed around the driver. Sure, there are vents and speakers in the back for passengers, but the positioning of the instrument panel is a clear indication that the driver has been the focus.
The steering wheel gets a lot of attention, but there are a lot of other features for which we’re listening to our audience and thinking about their needs, such as the location of the screen and the type of information riders can access. Waymo not only has the most on-road experience, we also have the most experience with customers. We have their survey feedback, which gives us a good amount of intel as we innovate. For example, we were hearing repeatedly from our riders that they want more variety in music options, so in addition to our iHeartRadio partnership, we recently integrated Spotify into our product to help meet that need.
The safety case for AVs
Emily Shao: You’ve said that the safety case alone is sufficient for pursuing autonomous vehicles. What’s a specific incident or data point that reinforced this belief for you?
Saswat Panigrahi: In the United States, more than 40,000 people die on the roads per year. That’s hundreds every single day. If you look at the number globally, it’s more than a million. What’s causing these fatalities is largely preventable human error, which none of us should be comfortable accepting. Since the pandemic, fatalities have been increasing.
Now juxtapose that with the reality that Waymo’s AVs have 80 percent fewer injury-causing crashes compared with an average human driver—this is based on empirical data, not forecasts or simulations. Clearly, scaling our service with this level of safety would be great. In my experience, both as a rider and as a designer, Waymo has prevented many collisions that I, as a driver, simply could not have.
Emily Shao: Are the safety numbers in line with what you expected from your simulations?
Saswat Panigrahi: For years, our safety simulation has indicated that Waymo would be a lot safer than human drivers. But now we can see that in our empirical data because we have published our results for 70-million-plus miles. Overall, no matter which metric you look at—from injury-causing collisions, pedestrian injuries, airbag deployments—we are 70 percent, 80 percent, or even 90 percent safer than cars with human drivers.
The tech-strategy nexus
Emily Shao: Waymo has driven over 100 million miles fully autonomously and billions more miles in simulation. How has simulation data directly influenced a change in strategy?
Saswat Panigrahi: The majority of our discoveries and improvements have come directly from simulation experiences. Let me give you an example. Let’s say that in the real world, you saw a car speeding—not by ten miles per hour above the speed limit but by 60 to 70. I didn’t know this kind of thing happened, but in Phoenix, we observed a case where somebody was driving about 90 miles per hour in a 25-mile-per-hour zone.
We wouldn’t often encounter a situation like this in real life, but with simulation, you can vary the speeds of all agents. This gives you far greater predictive power of how your car and its software will perform in these situations. Once you have built your system to be resilient in these situations, you can inject other elements, such as rain, or test how the vehicle performs at night.
We have the scale to continually simulate thousands of cars, which is a huge advantage. We invest heavily in AI and machine learning to make these scenes realistic while preserving the difficulty. It’s often an overlooked part of building an autonomous driver.
Emily Shao: What’s an undercover hero feature in a Waymo vehicle, something people don’t even notice, but it makes a big difference in their experience?
Saswat Panigrahi: Many collisions happen not while the car is driving but when riders are about to exit. People instinctively open the door without checking for bicyclists on the left or right of the car. The cyclist, who wasn’t anticipating the door opening, smashes into the door. This is called “dooring” risk.
We’re uniquely positioned to know exactly when we have reached our destination. We know the door is about to open. And we have a 360-degree view of where the cyclists are, even when the car isn’t driving. So we’re able to warn the rider that a cyclist is approaching and make sure that everyone is safe.
Emily Shao: Which cutting-edge technologies will have the biggest impact on AVs in the next ten years? How will they change the user experience?
Saswat Panigrahi: AI has already played an incredible part in solving core safety issues. It still has an important role to play in terms of understanding user intent and preferences. Say you’re getting picked up on the side of a 45-mile-per-hour road. You really care about being picked up on the right side of the road, but if you are on a narrow residential street, you may not want the car to go all the way to the end of the cul-de-sac, turn around, and come back and pick you up. We’ve been working on these kinds of nuances with the help of AI.
The second area where AI plays a big role is in simulation. There’s a tremendous amount of opportunity for creating virtual worlds that increasingly get harder to navigate so that we can learn how to deal with difficulties in the simulated world before they happen in the real world. Simulations also help us recreate real events where the Waymo vehicle needed help, so we can generate better solutions than if we were just looking at data logs.
Adapting to new markets
Emily Shao: Waymo is now operating in several cities: Phoenix, San Francisco, LA, Austin, Atlanta, and more to come. How does each city’s distinct urban personality shape how you design and deploy the technology?
Saswat Panigrahi: Our first cities were Phoenix and San Francisco, which we see as two extremes of urban driving experiences. We went fully autonomous with external riders in 2019 in Phoenix, which has some dense areas like school zones, but largely, it’s a high-speed environment, with a lot of 45-mile-per-hour thoroughfares.
The streets in San Francisco are much narrower, especially when cars are parked on both sides. Lots more cyclists, lots more pedestrians. What we had hoped, which thankfully turned out to be true, is that once you can handle the high-speed maneuvers needed in Phoenix as well as the dense urban environment of San Francisco, you’ll be able to handle a lot more scenarios.
Many other cities will fall somewhere in the middle. Take Los Angeles as an example. Parts of Hollywood are dense like San Francisco, but other areas have high-speed traffic like Phoenix. We were hoping that solving these two markets would allow us to move a lot faster, and that’s exactly what happened. Los Angeles is rolling out really well, and we have seen the same bear out both in Austin and Atlanta, where we have also launched.
Emily Shao: Waymo started test operations in Tokyo in January 2025. How are you thinking about international expansion?
Saswat Panigrahi: We have started testing in Japan and will evaluate multiple additional international cities. Think about when you go to a new city or country. You don’t need to relearn the entire task of driving; you just need a little bit of reorientation. That’s also true of our technology. Domestic launches—from Los Angeles to Austin and Atlanta—are getting faster, and we’re generalizing well. That gives us confidence that when we go international, we’ll have an autonomous driver that’s fundamentally safe. And the fact that we have been investing in artificial intelligence for more than a decade gives us a foundationally capable driving technology that can generalize.
The collaborative AV ecosystem
Emily Shao: Let’s talk about the ecosystem. Waymo has the vehicles, but so many other factors come into play to deliver the service. How do they all work together?
Saswat Panigrahi: Our goal is to make trillions of miles safer. We can only accomplish that if there’s an entire AV ecosystem. We rely on the expertise of many, many players in pursuit of our goal. There are infrastructure providers, folks who understand charging really well or who can do maintenance on tires and vehicles better than anyone else. We’re already working with many of them in our current deployments. Now that people can experience Waymo themselves, we’re seeing a lot of attention from players who, only a couple of years ago, may not have seen autonomy as a priority.
Emily Shao: You’ve spoken before about collaborating with regulators. What’s a successful partnership or a dialogue with the regulator that has helped advance Waymo’s deployment?
Saswat Panigrahi: Waymo has collaborated with regulators as well as key stakeholders from the federal, state, and city levels from the beginning. Even before we were fully autonomous and we were just testing, we have always had a very transparent and direct engagement approach.
At the federal level, this has involved collaborating with NHTSA [National Highway Traffic Safety Administration] and being more transparent than they require. So far, we are the only company that has released detailed information about all our collisions, no matter the cause. At a state level, we have collaborated with the DMV [Department of Motor Vehicles] in California to get permits for San Francisco and Los Angeles. In every city where we’ve deployed, we’ve worked not only with political offices but also law enforcement officers and first responders, undertaking direct testing with their vehicles to make sure Waymo responds properly to them. This transparent, open engagement has helped usher in a market transition in attitudes toward AVs.
Emily Shao: Looking ahead five or ten years, how is autonomy present in the world around us? What do our streets look like?
Saswat Panigrahi: As more and more people choose these options, more of the city will become available. Today, we subsidize more of our parking than our housing in some of our markets. This option will also result in many second-order benefits, such as freeing up the city from more parking structures. And most important, in addition to safety, we’ll be able to provide accessible transport options to people who may not have them today.
