Waymo and Wayve launch robotaxis in London with opposing approaches

Waymo and Wayve launched robotaxis in London almost simultaneously. This is not just competition for a market; it is a historic clash between two opposing philosophies of building autonomous cars. Both companies are absolutely convinced that their approach will dominate the future of mobility, and London is where that conviction will be tested against reality.

Waymo's philosophy: sensors, maps and engineering precision

Waymo chose the path of traditional engineering. Each vehicle carries many high-tech sensors: lidars (laser 3D scanners), spatial radars, stereo cameras and high-resolution monocameras. This multi-sensor system sees the world with redundancy: if one sensor fails, the others keep working.

But what is truly unique about Waymo is its maps. The company builds and constantly updates hyper-precise 3D digital maps of every route and street. These maps record the position of every lane marking, every road sign, the height of the curb, the position of the traffic light: literally everything a car needs to know about its surroundings.

Before launching in a new city, Waymo sends teams of specialists who spend many months collecting and processing data. Human annotators manually check that the car understands road scenes correctly. It is expensive and slow, but safe.

The strategy has worked. Waymo operates safely in San Francisco, Phoenix, Los Angeles, San Diego, Austin, Las Vegas and other large American cities. But Waymo's approach has an Achilles' heel: it demands huge capital investment and time. Expanding into a new territory means a months-long marathon of data collection, annotation and validation, and only then the car's first drive on the streets.

Wayve bets on pure AI and global scalability

Wayve sees the future very differently and rejects the concept of detailed mapping. Instead, the company relies on computer vision and deep machine learning trained on a colossal volume of video data.

Wayve's cars are equipped with cameras, but the machine-learning task is fundamentally different. The system does not learn from the data of one specific route, but from the aggregated experience of hundreds and thousands of cars driving through different cities, in different conditions, in different seasons and at different times of day. The neural network learns to recognize typical road scenes, the behavior of pedestrians and cyclists, and the actions of other drivers, not tied to a specific geolocation but as universal patterns that transfer between cities.

Wayve claims its system already works in London and other European cities with minimal advance preparation. The company says it can expand much faster, because it does not need a months-long marathon of map collection and annotation before each launch. A new city is simply new video data feeding the existing neural network.

London: the historic meeting point of two worlds

London is the first city where both approaches have launched at the same time in real commercial conditions. This is no coincidence; it is a chance to see which philosophy works better, which is safer, which is more economical.

The test will last months and years. Observers will track safety, reliability, speed of expansion and operating costs. The industry is frozen in anticipation:

• Which method will deliver the best safety in real London conditions?
• Can Wayve's system, without detailed maps, function reliably in a new city?
• Which approach is cheaper when scaling to tens and hundreds of cities?
• Who will reach operational profitability first?
• Which approach will win more investment and consumer trust?

What this means for the future

The outcome of the London test will shape the architecture of global mobility for a decade. If Wayve is right, robotaxis will be available in all major cities of the world much sooner and significantly cheaper than experts currently assume. That means the era of human driving will end faster than many think.

If Waymo is right, investment in high-precision mapping and multi-sensor systems will pay off, but global expansion will stay slow, capital-intensive and concentrated in a few large cities. In that scenario, robotaxis become a premium-segment product, available only in developed countries.

London has become a natural proving ground for two visions of the future. Over the next few years, the whole world will watch the results.