Intelligence on machines that act
Sector
Cameras, sensors and perception software giving machines the ability to see, inspect and understand the physical world.
One measures the world; the other infers it. A decade-long fight over how machines should see has quietly ended in a truce — and the terms of that truce explain most of what's on the roof of a robotaxi.
Driver-optional farm machines are working fields at Level 4 today, years ahead of urban robotaxis. The reason isn't better technology — it's a fundamentally easier problem, and understanding why explains the whole sector.
Transport for London has licensed Wayve's camera-only, mapless fleet for Uber rides — putting a homegrown AV champion ahead of Waymo in the capital, and staking the UK's autonomy debut on the riskier of the sector's two philosophies.
The least glamorous autonomy — the camera watching a production line — is having its AI moment. Edge processors, example-trained models and a shift from "detect the defect" to "judge the part" are quietly rewriting factory economics.
Zero-shot inspection models are replacing hand-tuned vision pipelines that took decades to build. Here's what changes, what doesn't, and what to ask your vendor.
The Autonomy Briefing
News, deals, deployments and regulation — curated by our editors every Thursday. Written for engineers, executives and policymakers. One sponsor per issue, clearly labelled.
