Toyota and Turing advance autonomous driving via proprietary E2E AI

Automotive companies are advancing end-to-end (E2E) autonomous driving technologies through proprietary AI and specialized hardware. Toyota Motor Corporation is developing Level 2++ autonomous driving features, aiming for market introduction by 2028. This system utilizes a unique AI paired with a rule-based “guardrail” mechanism designed to intervene if the AI makes unexpected decisions, serving as a safety layer and providing explainability in the event of an incident. Toyota’s approach is integrated via its Arene software platform, which aims to provide affordable E2E capabilities across various vehicle segments.

In a similar pursuit of advanced automation, the Japanese startup Turing is working toward Level 5 full autonomy. Unlike traditional methods that rely on heavy sensor suites including LiDAR and radar, Turing utilizes a camera-centric approach. By equipping modified commercial vehicles, such as the Toyota Alphard, with only seven cameras and high-performance onboard computing, the company aims to achieve smooth urban driving through superior AI performance. Turing has successfully conducted test runs in several Japanese cities, including Tokyo and Osaka, demonstrating the ability of its E2E AI to navigate unfamiliar environments using data-centric computing infrastructures.