Humans have been obsessed with self-driving cars since time immemorial. In films, the fascination dates to 1927. Metropolis, a silent feature deemed both silly and the first of its kind science fiction epic, imagined a futuristic city where sleek, automated vehicles traversed along elevated highways in a highly coordinated fashion. As they say, if you can imagine it then you can make it and so, humankind too, has been on a mission to make them ever since. It was not easy. It took a lot of time. But finally, we seem to be getting there.
Earlier in the year, when Alphabet and Google CEO Sundar Pichai took the stage of I/O, to give the developer community a sneak-peek into all the AI his company was preparing to unleash on the world, he left the keynote with one very personal anecdote. He told everybody about the time when his parents were visiting him in San Francisco. The first thing they wanted to do was to ride in a Waymo. Pichai had obviously used Waymo taxis before but watching his father, who is in his 80s, being totally amazed by self-driving technology made him see the progress in a whole new light, he admitted.
If you can read between the lines, this was all the assurance that anybody needed, that the “future” has officially moved from the laboratory to the curb. Not just in San Francisco, but even if you live in Beijing, Phoenix, or Abu Dhabi, the sight of a steering wheel spinning by itself is no longer a reason to be spooked and call the cops: it is just how people get to brunch today.
Who is winning the race?
According to the RoboTaxi Global Market Report 2025, the global robotaxi market is growing exponentially, up from $1.19 billion in 2024 to a projected $2 billion-plus this year. The increasing penetration of autonomous technology and need for sustainable and eco-friendly travel is a big growth driver. With improvements, scale and affordability pitching in eventually, the market size is expected to grow further, reaching tens or even hundreds of billions of dollars by 2030.
Waymo clocked over 20 million miles and 14 million paid rides in 2025 following concerted expansion across Phoenix, Los Angeles, and San Francisco. Meanwhile, Zoox which is owned by Amazon, continues to ramp up testing for its unique “toaster-shaped” bidirectional pods that lack a steering wheel and can travel up to 75 miles per hour.
In China, Baidu’s Apollo Go dominates at an unprecedented scale in cities like Wuhan and Beijing. As of October 31, 2025, it recorded over 250,000 weekly fully driverless rides, on par with what Waymo reported in April for its equivalent service in the US. Despite the speedy ramp up, Apollo Go robotaxis have not been involved in any major accident involving human injury or death, Baidu claims. The Chinese government itself is helping expedite the adoption by building “smart roads” with sensors to talk to these robotaxis.
Looking at all this, Europe is finally waking up as well. Waymo recently announced a London launch for 2026, and Germany has become a primary testing ground for SAE Level 4 autonomous driving backed by some of the world’s most robust regulatory frameworks thanks to the Act on Autonomous Driving.
Behind the scenes, there are two very different types of technologies at work, now. The likes of Waymo and Baidu are following the modular approach where the robotaxis use a combination of LiDAR, radar, and high-definition (HD) maps, so technically they know every curb and traffic light before they even start the engine. It is supposedly super safe, but it is also expensive to run and scale while being increasingly prone to confusion and possibly malfunction if the map is updated suddenly, but its data is not. To counter this potential loophole and to lower the cost, Elon Musk has mandated a “vision-only” approach at Tesla, omitting maps and lasers completely, instead using AI to see and navigate the world. The catch is that without LiDAR, the Autopilot and Full Self-Driving (FSD) systems lack a fail-safe for when the cameras are blinded by glare or heavy rain.
The bumps in the road
Despite the 68 percent annual growth rate projected in the RoboTaxi Global Market Report 2025, the road to a fully self-driving future isn’t without its share of roadblocks. Evidently, even the smartest AI can be outsmarted by a simple power outage. This was the case after a localised grid failure froze a fleet of modular robotaxis recently in San Francisco. Because the traffic lights went dark and the HD maps no longer matched the intersection on the ground, the robotaxis defaulted to safe mode, basically turning into shiny bricks of metal with nowhere to go and causing a massive traffic jam.
The streets are full of elements of different kinds. Some people like to fool around and robotaxis – as luck would have it – are easy targets. As it turns out, you can confuse the sensors and force the car to stop simply by placing orange traffic cones on its hoods. There are also deep concerns about machine uprising and job displacement. With 2.5 million robotaxis expected to be on the road by 2030, human taxi and Uber drivers are understandably worried about their livelihoods.
Then there are the edge cases. How does a robotaxi react to a construction worker using hand signals? Or a child dressed as a traffic cone for a costume party? The real-world dynamics are so complex, it would take an incredibly high number of tests to validate safety, far exceeding anything that exists for human drivers today.
Why we still want them
If technology is so hard, the question is, why do we still want to do this? Driving can be painful and excruciating, especially in bustling cities with breakneck traffic, but it can also be remarkably freeing – so why do we want to replace that with sensors and cameras and AI? There are no right answers. Also, there are no wrong answers. It is a choice that – given how fast the technology is moving – we will have to make at some point.
What we know is that self-driving cars have certain very promising advantages. Per a McKinsey study, the average commuter could save up to 50 minutes a day using them, a time they could spend doing something else like taking up a hobby or just getting some much-needed rest after a hard day’s work. Plus, all the space you save inside the car can be put to other use. It might even open new avenues of business for co-working spaces and movie theatre chains.
For the visually impaired and elderly, robotaxis can be a ticket to independence. Despite the headlines about accidents, the data is becoming clear: robots don’t get drunk, they don’t text while driving, and they don’t get road rage because someone cut them off.
As the world ushers in the new year, we are also moving past the era of “if” and into the era of “where” for robotaxis. Technology has proven it can handle the streets. The next bumps aren’t going to be technological, they will be legal and social. Who is liable in a crash? How do we tax robot miles to fix the roads?
The driverless world is no longer a dream. It’s a real service that many around the world can download on the App Store. As Pichai said in his final remarks: It is a reminder of the incredible power of technology to inspire, to awe and to move us forward.
