The technology exists. The money doesn’t.
I’ve been tracking the London Underground automation debate. Driverless trains work in dozens of cities worldwide. London could build them tomorrow.
But £20 billion changes the conversation entirely.
That’s Transport for London’s February 2025 projection to convert three of its oldest lines to driverless operation.
The Age Tax
London operates the world’s oldest metro system. Parts of the infrastructure are over 130 years old, built deep underground with single-bore tunnels and no space for side walkways.
Full automation requires platform edge doors on every station, camera systems, new signaling infrastructure, and reengineered stations. Retrofitting Victorian-era tunnels multiplies every cost.
Paris converted Metro Line 4 to fully automatic operation for €470 million—roughly £400 million.
London’s projected cost for three lines: fifty times higher. The difference is infrastructure age.
When Math Kills Dreams
Operational savings reach maybe £100 million annually in reduced wage costs. At that rate, recouping a £20 billion investment takes two centuries.
Union representatives call it a “political myth.” The numbers prove them right.
The Decision
In 2024, Mayor Sadiq Khan abandoned plans to introduce driverless trains. The economics didn’t work.
What Legacy Costs
Dubai runs the world’s longest driverless metro line. Singapore and Paris operate automated networks. The technology works everywhere except where it matters most—in systems built before the technology existed.
London’s 160-year-old infrastructure creates transition costs that newer systems never face. Every tunnel, platform, and station compounds the bill.
Twenty billion pounds could build an entirely new metro system. Instead, it buys three automated lines in an old one.
That’s the real cost of legacy infrastructure: not maintaining it, but changing it.