We live in an automotive world. No matter how good (or more frequently bad) your country's public trains and trams are, the fact is that most of the planet is propelled by the motor car. 

For most of the history of the car, if you couldn't or wouldn't drive yourself around, you had to employ another human – either a chauffeur or a taxi driver – to do so, continuing a pattern that stretches back through horse-drawn hackney cabs and hansom carriages to human-hefted rickshaws. 

But now, we're entering the era of the autonomous automobile – a driverless vehicle, navigating using AI and a sensory perception system that owes more than a little to… bats.

This is Future Mode, and this is autonomous driving. 

How robotaxis know where to go 

We join CGTN's Wang Tianyu in what, just decades ago, would seem like a nightmare scenario: zooming along city streets in a car without a driver. 

Tianyu's fully autonomous 'robotaxi' doesn't need a driver to check the mirrors, pump the pedals, spin the wheel and work out the best routes - not in the age of superfast technology. 

The purely mechanical movements (stop, go, turn) come at the behest of AI, the driver's brain – and it's a brain constantly updated with data feeds calculating possible routes. 

That's a feedback system you may already be aware of and benefitting from, if you've ever taken advice from your favored online-mapping app that "a faster route is available". However, there's a feedback system on robotaxis that takes awareness to the next level. 

The car is dotted with cameras, observing the surrounding traffic from every angle, but that information is supercharged by an idea borrowed from bats.  

The winged mammals are famed (and feared) for their ability to fly at night without bumping into anything. They achieve this by echolocation – emitting noise, listening to how it bounces back off objects and prey, and using that input to calculate a moving 3D map of their world.

Inspired by the bats and terrified by the Titanic disaster, humans adapted echolocation into sonar (SOund Navigation And Ranging) for use in boats and submarines; many cars' parking sensors use a similar system. 

However, sound is impacted by atmospheric conditions; in poor weather even bats, evolved over millions of years, can only detect up to two meters away – far too close for vehicles traveling at speed on public highways.

Instead, robotaxis use lidar (LIght Detection And Ranging), bouncing pulses of infrared laser light off their surroundings in order to build up a 3D map of the environment. 

The information from these 'extra eyes' feed into the car's 'digital brain', allowing it to track movements and react in real tim – and in the back of his rush-hour taxi, Tianyu can relax. 

And he's not the only one. In China, Pony.AI and Baidu's Apollo are among the major players, each of them operating more than 1,000 robotaxis on public roads. As bookable as regular human-driven cabs, they're changing commuting patterns and improving safety and efficiency in urban environments. 

As Tianyu says, "I feel like the car knows every single road and alley, so it can make decisions in advance. But it can also respond to the unexpected – hitting the brakes in the blink of an eye."

Europe gets on the road to robotaxis 

If China is at the vanguard of robotaxis, could Europe also be a home to this transport revolution? CGTN's Michael Marillier took to the streets of London to canvas public opinion. 

"I see it all over social media – I definitely would get in one, for sure," enthused Misha. "It would be quite fun to have self-driving cars in London."

"Cars coming out of China are exceptional with the technology that's built into them - and incredibly affordable," said Bowha. "So I've got more trust now than I did, as short as two years ago."

One slightly older respondent, Chris, was a touch more hesitant – but wouldn't rule it out: "I think maybe for those driving for a significant amount of years, it may take a little bit longer for us to get used to the concept."

Those members of the public could ask their interviewer for his own experience of Chinese robotaxis. In Shanghai recently, Michael jumped in a Pony.ai robotaxi and found the experience, in his words, "kinda spooky."

"There's no one in front of me," he reported from the back seat. "It's like we're being driven by a ghost. But here we are – this is the future."

And it's a future that could be heading west – Pony.ai is hoping to offer robotaxi services in European cities. But that's not as simple as it sounds: Europe's roads are different to those in China, and its laws are different too, not to mention differing between countries.

And that's why the company is setting up shop in Luxembourg – nestled usefully between France, Belgium and Germany. It's running a test program, training its systems to understand Europe, one trip at a time.  

"The key thing is figuring out – how do people drive differently? How do they react differently in different situations?" asks Andreas Reschkha, Pony.ai's Director of Product, Systems & Safety. 

Safety and testing 

So the wheels are turning in Europe, but how long will that journey take? 

In China, the industry has made steady progress, moving from the test phase to daily use. It took time, but the robotaxi revolution accelerated in late 2022 when officials decided that some services no longer required a safety driver.

"We actually have multi-layers of redundancy," explains Pony.ai founder and CEO James Peng. "All those hardwares and softwares, and AI design, has intertwined and integrated together to ensure safety. 

"Our record has shown, it's almost 10 times safer than a typical human driver in our operating areas."

Pony.ai, which clocked up more than 60 million kilometers during test drives, currently generates almost half of its revenue from another mobility business: self-driving trucks. 

These vehicles travel along highways and major logistics routes – helping companies move goods between cities. And analysts say self-driving services can build on their experience in China.

However, it may not be immediate. Ron Zheng, of consultancy firm Roland Berger, notes that Chinese firms will have to tweak the technology to Europe's roads, with their baked-in complexity.

"Chinese autonomous driving companies need to retrain the models," Zheng tells CGTN. "European roads feature narrow streets, roundabouts are very common and there are also mixed urban, suburban and rural traffic conditions."

Back in Luxembourg – and back at the testing stage – it's time to hit the road. The safety driver holds the steering wheel, but only steers in an emergency. 

Regulators could allow the cars to operate without backup drivers if they pass a series of tests. That may take several years, but the company says artificial intelligence is accelerating the learning process. 

"We use a lot of generative AI – especially in our simulation environment – to create artificial worlds, to create more test cases," says Reschka. "And to test things that we can't experience on the road, but we know we will have to solve when we experience it."

The road to the future can be a slow crawl, or a sudden dash. But this is a route that's already been traveled – and the signs are good that we'll reach our destination soon.

Click here to enjoy our full interactive Future Mode special