Auto Spring, the World’s First Inland Test Vessel for End-to-End Autonomous Navigation, Successfully Launched

On 10 April, Auto Spring, the world’s first clean-sheet inland test vessel designed for end-to-end autonomous navigation, was successfully launched in Huzhou, Zhejiang Province.

Auto Spring has an overall length of 65 m, a container capacity of 64 TEU, fully electric propulsion, and a range of more than 200 km. The vessel is scheduled for deployment on inland waterways in northern Zhejiang.

Designed for scalable commercial adoption, Auto Spring is a clean-sheet, next-generation smart vessel that represents a new class of inland vessel.

Auto Spring is the first vessel to introduce the automotive concept of electrical/electronic architecture, or E/E architecture, into the marine sector. Its next-generation central domain platform enables efficient data connectivity and unified resource coordination. The E/E architecture significantly reduces electrical system costs while closely integrating the propulsion and autonomous navigation systems, providing a robust hardware foundation for maritime autonomy.

Electric Propulsion

Drawing on the strengths of the electric vehicle supply chain and adapting them to inland operating conditions, the project developed the industry’s first marine 800 V high-voltage DC electric propulsion system. The system has received type approval from a classification society.

Compared with conventional marine electric propulsion systems, the solution improves energy efficiency while significantly reducing costs through the integration of automotive industry technologies and supply-chain capabilities. It provides an important practical foundation and commercial pathway for the large-scale electrification of inland vessels.

Auto Spring also adopts a cycloidal propeller. Compared with conventional screw propellers, it provides higher propulsive efficiency, is better suited to shallow inland waters, and offers 360° thrust vectoring and maneuverability, creating favorable conditions for autonomous navigation.

Autonomous Navigation

The perception architecture is built on a BEV-based multitask deep neural network that fuses data from LiDAR, millimeter-wave radar, cameras, and other sensors. A unified deep neural network performs multiple perception tasks and supports adaptation across different weather and operating conditions.

The system is designed to support autonomous navigation, collision avoidance, overtaking, berthing, and unberthing across a wide range of scenarios, with the goal of achieving berth-to-berth autonomous navigation on inland waterways.

Auto Spring also features a complete closed-loop data framework. Data collection, model training, simulation, and OTA updates are integrated through a cloud platform that combines onboard intelligence with cloud-based capabilities. This creates a closed loop for algorithm training and supports rapid development and iteration.

The Auto Spring project was initiated by Han Bin, founder of ShipFinder and a serial entrepreneur. Through the establishment of BOTIX, the project draws on China’s strong industrial and talent base in electric vehicles and autonomous driving to advance the electrification and intelligent transformation of inland shipping in China and worldwide.

Following its launch, Auto Spring will undergo mooring trials and autonomous navigation trials. The project will progressively advance the deployment of China’s first autonomous navigation route, continue exploring scalable adoption pathways for green and intelligent vessels, and accelerate the commercial development of maritime autonomy.