China’s WeRide unveils Bosch-backed AI ADAS for 2025 launch

WeRide has introduced WePilot AiDrive, a one-stage end-to-end advanced driver assistance system (ADAS) developed with Bosch.

WeRide, listed on Nasdaq and based in China, specializes in autonomous driving technology.

The new system combines sensing and decision-making into a single process, which the company says could allow for faster responses and improved fault tolerance compared to traditional two-stage ADAS approaches.

WeRide said WePilot AiDrive has completed core function validation and is scheduled for mass production and vehicle deployment later in 2025.

The company claims the system can handle scenarios such as lane changes in heavy traffic, detours around construction, unprotected turns, and interactions with pedestrians and obstacles.

WeRide reports that the system operates on multiple computing platforms and adapts to different sensor setups.

Testing has included simulated scenarios, and the company has showcased a test vehicle in Guangzhou.

No independent performance data or deployment partners have been disclosed.

🔗 Source: WeRide

🧠 Food for thought

1️⃣ Realistic timelines emerge after years of overly ambitious predictions

WeRide’s 2025 target for mass production reflects the industry’s shift toward more achievable timelines after learning from earlier miscalculations.

Major automakers consistently overestimated their capabilities in the late 2010s, with General Motors aiming to mass-produce fully autonomous vehicles by 2020 and Ford targeting Level 4 vehicles by 20211.

Both companies faced significant delays due to safety testing requirements and technical complexity that proved more challenging than initially anticipated.

The six-month gap between WeRide’s two-stage solution reaching mass production and now unveiling their one-stage system demonstrates the rapid iteration cycles that characterize today’s more methodical approach to autonomous vehicle development.

This measured progression contrasts with the overly ambitious mentalities that dominated earlier industry projections, suggesting companies now better understand the engineering realities of bringing autonomous systems to market.