Physix Frontier
How to Call London Uber Autonomous Cars: Beginners Start Here
Over the weekend in London, I took an Uber autonomous ride and stepped into quite a few pitfalls. My judgment is: worth experiencing, but not suitable for commuting. It's more like a city-level gray-scale test, opening in limited areas on a small scale first, not something you can hail anywhere in the city. Whether the car reaches your downstairs depends on maps, operational zones, vehicle dispatch, and regulatory boundaries, regardless of your willingness.
For the previous month, I frequently used L2 assisted driving, where the car helps maintain lane and following distance, but humans must watch the road. This time, seeing BBC reports, London launched the UK's first wave of robotaxis, i.e., self-driving ride-hailing cars, provided by autonomous solutions like Uber and Wayve. Behind this is a batch of modified cars integrated into ride-hailing platforms; ordinary cars with a switch added can't do this.
Reports mention that Uber says it's available now, but the UK government places the approval time for fully autonomous vehicles in 2027.
Beginners tend to oversimplify selecting car types in the app. Opening the app doesn't guarantee hailing a driverless car; you must first be in designated London areas, and the app must show the car type. There is usually a human driver or safety operator throughout, responsible for monitoring the road and taking over. Regarding price, I didn't see a stable conclusion; it might just be standard billing. It's more suitable for short-distance experiences, with limited routes and pickup points, not for daily commuting.
First hail: Confirm you are in London. Open Uber, drag the map to the city center or your residence. If the app doesn't show vehicles, don't panic; your location might not be in the operational zone. When entering origin and destination, pick a clear roadside point for origin, not "current location." Pick popular landmarks like museums, stations, or park entrances for destinations to see available car types more easily. Look for autonomous taxis in the car type list; names might be robotaxi, autonomous car, or just "Autonomous Vehicle." No option means no coverage currently, time slot not open, or vehicles all taken.
Confirming pickup points and payment is where beginners stumble most. The app selects "parkable spots," which aren't necessarily at your door. Change to a curb-side intersection where parking is allowed. While waiting, verify license plate and color. When the car arrives, don't rush to pull the door; confirm it's dispatched by the platform, then check if anyone is in the front row. Drivers or safety operators won't ask you to change destinations. Sit in the back seat and buckle up. Don't touch the steering wheel or screens during the trip. If the car changes lanes, turns, or stops, observe if it's smooth. The car drives itself; humans only supervise. Save trip records after getting off. Reviews can say "experienced once," not "fully unmanned." Platforms consider real demand and failure rates when judging future area expansion.
Pitfalls separately:
Location drift is the first pitfall. London has many underground structures and tall buildings; GPS might throw you across the street. Manually drag the pickup point to the roadside; don't trust "current location."
Thinking the car is truly empty is the second pitfall. Currently, these services likely still have human drivers or safety operators. Encountering construction, jaywalkers, or complex intersections, humans might take over instantly. Sitting in the back, don't be nervous, but don't treat this as fully autonomous driving.
Payment failure is the third pitfall. If you use a local bank card or haven't bound a card, travel in London might get stuck. Bind a usable payment method in the app before departure; amount doesn't matter, as long as it can deduct.
Complex routes are the fourth pitfall. Don't choose airport-to-city-center long distances for the first time. In my tests, short trips made it easier to see how it handles normal intersections. Long routes pass through more construction zones, bus lanes, and temporary controls, increasing failure probability.
Treating experience as conclusion is the fifth pitfall. The real difficulty of autonomous driving is cramming rain, pedestrians, bicycles, taxis, and buses onto the same road every day. Demos are just the beginning. I heard about data flywheels when using recommendation systems; cars are similar—the more they drive, the more abnormal scenarios they encounter, and the more stable the model becomes.
If you happen to be in London, spend one short trip experiencing it, treating it as an opportunity to observe the boundaries of autonomous driving. Afterwards, try hailing two regular Ubers to the same destination, recording wait times, pickup points, and price differences to see if autonomous cars really save effort.
📌 This article is compiled from BBC Tech, original text https://www.bbc.co.uk/news/videos/cwyzeyd7peeo?at_medium=RSS&at_campaign=rss
Copyright belongs to the original author; this is a compilation and independent analysis based on public reports.