Lenovo and South Korean autonomous mobility firm SWM have announced a partnership to co-develop a Level 4 autonomous robotaxi platform designed for deployment in several international markets.
The platform will combine SWM’s AP-700 Autonomous Driving Platform with Lenovo Vehicle Computing’s AD1 Autonomous Driving Domain Controller, which is built on the NVIDIA DRIVE AGX Thor computing platform.
According to the companies, the AD1 system is designed to support high-performance autonomous driving applications and is based on NVIDIA Blackwell Architecture, delivering more than 2,000 TFLOPS of artificial intelligence processing power at FP8 and INT8 precision.
The computing platform is intended to support complex workloads such as transformer-based models and generative AI applications used in autonomous vehicle perception and decision-making.
SWM said the collaboration will allow the AP-700 platform to integrate advanced computing capabilities needed for Level 4 autonomous driving in dense urban environments.
The company has been developing autonomous driving technology since 2017 and launched South Korea’s first fully driverless taxi service in the Gangnam District of Seoul in September 2024.
The robotaxi service has been operating without a safety driver and has maintained a clean safety record since its launch, according to the company.
SWM works with several public-sector partners in South Korea, including the Seoul Metropolitan Government, the Seoul Taxi Association, Gwangju Metropolitan City and the Ministry of Trade Industry and Resources.
The new robotaxi platform is expected to be integrated into the NVIDIA DRIVE Hyperion ecosystem, which combines AI computing hardware, sensor architecture and software designed for autonomous vehicles.
Initial commercial deployments are planned in Japan, markets across Southeast Asia and parts of the Middle East, with the companies aiming to expand robotaxi services to additional global markets over time.
Peter Xu said the AD1 controller extends Lenovo’s hybrid AI strategy into the automotive sector by providing a vehicle-grade computing platform designed for safety and performance.
Ki Hyuk Kim said the partnership will combine Lenovo’s AI computing infrastructure with SWM’s autonomous driving software to support the global expansion of robotaxi services.
Rishi Dhall said the collaboration demonstrates how developers can move autonomous driving applications from development into real-world deployment using a unified computing architecture.