DeepRoute.ai has introduced its AI Spark platform, expanding its artificial intelligence technology beyond self-driving cars. The new platform is designed to help various moving objects, not just vehicles, understand and navigate their surroundings using AI. This is part of the company’s broader goal to create RoadAGI, a system that applies artificial intelligence to roads and mobility.
AI Spark uses a Vision-Language-Navigation model, allowing objects to follow spoken instructions, recognize their environment, and travel autonomously without depending on high-definition maps. The system is designed to work both indoors and outdoors, ensuring safe and flexible movement in different environments.
This development builds on DeepRoute.ai’s success in producing smart driving technology. Since August 2024, over 40,000 vehicles with the company’s self-driving system have been sold, and newer models featuring an improved AI model will be launched later this year. DeepRoute.ai expects to surpass 200,000 vehicle sales as it continues improving its AI-powered navigation and perception technology.
CEO Maxwell Zhou said that expanding beyond self-driving cars is a natural next step for the company. By using millions of miles of road data, DeepRoute.ai aims to develop a “road brain” that can power different types of moving objects. Zhou explained that AI Spark could enable robots and other automated machines to move smoothly from indoor spaces to public roads and parks.
DeepRoute.ai is showcasing AI Spark and its RoadAGI strategy at the NVIDIA GTC event at booth #1843. Zhou is also giving a virtual presentation about the company’s plans to expand AI-driven mobility beyond cars.
