Toyota Motor Corporation and its software subsidiary, Woven by Toyota, officially unveiled a new suite of artificial intelligence technologies on April 22, 2026, intended to serve as the digital backbone for Woven City. The centerpiece of the announcement is the AI Vision Engine, a large-scale AI foundation model designed to integrate urban infrastructure with autonomous mobility and personal robotics.
The AI Vision Engine utilizes a multimodal architecture capable of processing high-resolution video feeds from thousands of distributed sensors located throughout the Woven City site in Susono, Japan. According to technical specifications released by Woven by Toyota, the model is trained on vast datasets of urban interactions, allowing it to recognize complex patterns and predict potential safety hazards before they occur. The engine operates with a latency of less than 50 milliseconds, enabling real-time intervention for autonomous vehicles and support robots operating in shared spaces.
A key feature of the AI Vision Engine is its co-creation capability. Toyota representatives stated that the platform provides a standardized API for third-party developers and partner companies to build and test new services. This includes automated logistics systems, personalized healthcare monitoring, and energy management solutions. By providing a unified vision layer, the engine allows different robotic systems to share a common understanding of the environment, reducing the computational load on individual devices.
During the presentation, Woven by Toyota executives highlighted the model's role in enhancing the Arene operating system, Toyota’s proprietary software platform for programmable vehicles. The AI Vision Engine serves as the perception layer for Arene, facilitating seamless transitions between manual and autonomous modes while ensuring compliance with local safety protocols. The deployment marks a shift from discrete AI applications to a centralized foundation model approach for urban management.
The announcement also detailed the initial deployment scale within Woven City’s Phase 1 area. The system currently manages data from over 2,500 integrated sensor nodes and supports a fleet of 150 autonomous mobility-as-a-service units. Toyota confirmed that the AI Vision Engine would be iteratively updated using federated learning techniques to protect resident privacy while improving model accuracy.
Official statements from Toyota emphasized that these technologies are not limited to the laboratory environment of Woven City. The company plans to begin exporting specific modules of the AI Vision Engine to its global vehicle lineup starting in the 2027 model year. This rollout aims to standardize advanced driver assistance systems across different regions and vehicle segments.