Automotive AI Software News: TIER IV Releases E2E Models

TIER IV Releases Multiple Open-Source End-to-End Autonomous Driving AI Models via Autoware

TOKYO, Japan - November 11, 2026 -TIER IV, the developer behind the open-source Autoware autonomous driving platform, has released multiple open-source end-to-end (E2E) AI models for autonomous driving. The release is relevant to the Automotive Artificial Intelligence (AI) Software Market, as automakers and suppliers look for AI models they can adapt to their own vehicle designs.

The models support different system requirements, with options ranging from camera-only setups to configurations that add LiDAR, radar and high-definition (HD) maps. Autoware provides the development environment to optimize the models and integrate them into vehicle systems. The release builds on Autoware 2.0, an architecture that combines AI models with conventional rule-based technologies.

The release covers five models: NVIDIA Alpamayo 2 Edge, AutoE2E, METEOR, OnePlanner and LatentDrive. They differ in size and hardware support. METEOR has 16 million parameters and works across semiconductor platforms, including NVIDIA and Renesas. AutoE2E has 79.9 million parameters and can run on computing platforms beyond NVIDIA hardware. Alpamayo 2 Edge has 2 billion parameters, compared with 34 billion in Alpamayo 2 Super.

According to the Next Move Strategy Consulting report, the global automotive AI market, which covers hardware and software components, was valued at USD 4.29 billion in 2023. It is forecast to reach USD 25.78 billion by 2030, a CAGR of 29.2% from 2024 to 2030.

TIER IV founder and CEO Shinpei Kato said automakers and component suppliers need to be able to adapt AI technologies to their own applications without depending on a single AI model or semiconductor platform.

Key Highlights:

• The release spans camera-only and multi-sensor configurations, including LiDAR, radar and HD maps.

• Hardware flexibility varies by model: METEOR supports NVIDIA and Renesas platforms, and AutoE2E runs beyond NVIDIA hardware.

• Model sizes range from 16 million to 2 billion parameters, giving developers options suited to different in-vehicle performance requirements.

• Integration with Autoware is complete for Alpamayo 2 Edge, the AutoE2E Reactive Model and OnePlanner. Integration and in-vehicle validation for LatentDrive are under way.

Analyst Insight:

According to analysts at Next Move Strategy Consulting, models that span camera-only and multi-sensor configurations are consistent with the advances in machine learning and computer vision that the firm identifies as a key catalyst for automotive AI adoption. NMSC analysts note that portability across semiconductor platforms is relevant to the high development and integration costs that its research lists among the restraints on adoption.

Industry Outlook:

TIER IV has said it will continue improving the models’ performance and optimizing them for in-vehicle environments. It will also support commercialization by automakers and partners through Autoware. Development of AutoE2E’s World Model and Reasoning Model is continuing, and TIER IV is considering commercialization for robotaxi applications. How quickly these open-source models move from release to in-vehicle validation will indicate the pace of adoption. NMSC’s forecast points to continued expansion of the automotive AI market through 2030.

Source: PR Newswire

For More Information: Download FREE Sample on Automotive Artificial Intelligence (AI) Software Market Report