Chinese artificial-intelligence startup Jijia Shijie has signed a strategic framework agreement with Changjiang Automotive Electronics to bring embodied intelligence to automotive electronics manufacturing, according to Chinese outlet D1EV.
The two companies plan to build what they call a data flywheel that runs from real production scenarios through data collection, model training, deployment, feedback and back to collection, with Changjiang's actual factory floor as the test bed. The aim is to validate embodied AI - machines that perceive, reason and act in the physical world - in a real automotive electronics plant.
Three-layer cooperation
The agreement sets out three layers of cooperation. First, the pair will jointly assess the feasibility of industrial data collection, labelling and data-product development, building a data foundation tailored to automotive electronics manufacturing. Second, Jijia Shijie will contribute its world model, embodied foundation model and software platform, which will be trained and deployed at real workstations using Changjiang's process, quality and equipment conditions. Third, once the first scenario is proven, the partners will explore replicating the solution across Wenzhou's auto parts and motorcycle parts supply chain and to other manufacturers, and discuss extending cooperation to computing resources, hardware manufacturing, robot bodies, component supply and contract manufacturing.
Changjiang Automotive Electronics says its business covers 10 countries, serving more than 60 OEM customers and participating in the development of more than 500 vehicle models. The company claims one in every three cars on the road carries its products.
In the embodied intelligence field, Changjiang is already working with the city of Wenzhou to build a robot training and scenario verification centre, according to the report.
Context
The deal reflects a broader push by Chinese EV suppliers to apply embodied AI - robots and software that can operate in real production environments - to manufacturing, as automakers and parts makers seek to cut costs and improve quality. The automotive electronics sector, which includes components such as controllers, sensors and wiring, is seen as a promising early use case because tasks are repetitive and environments are controlled.
Jijia Shijie is a relative newcomer in the AI software space, and the agreement with Changjiang is an early attempt to commercialise its world model and embodied foundation model outside the lab. No financial details or timeline were disclosed in the announcement.
