Beijing (Gasgoo)- On March 25, Ji Yue, the automotive robot (Robocar) brand jointly developed by Baidu and Geely, hosted the AI DAY 2024 event, where it announced the release of the OTA V1.4.0 software upgrade. This upgrade encompasses various domains, including intelligent driving, smart cabin, intelligent connectivity, and electrification.
The V1.4.0 version marks Ji Yue's third consecutive OTA release within 90 days and the fifth overall since its market debut. With over 200 upgraded features, the new software version addresses more than 5,300 user feedback issues.
During the AI DAY event, Baidu presented three of its latest AI native technologies in the field of autonomous driving, LD (Lane Definition) map, and human-machine interaction. These cutting-edge AI technologies will be integrated into Ji Yue’s Robocars.
At the event, Baidu officially introduced the Apollo Vision Takes All (VTA) autonomous driving visual perception large model. The VTA significantly enhances capabilities such as dynamic and static object detection, temporal tracking, real-time mapping, and scenario understanding. This upgrade further elevates Ji Yue's advanced pure vision-based autonomous driving capabilities.
Regarding intelligent driving maps, the discussion around with or without HD map has been a hot topic in the realm of autonomous driving. Baidu's LD maps, which provide lane-level navigation, are native maps generated by their autonomous driving visual perception model. This innovative approach revolutionizes traditional map production methods and overcomes the scalability limitations of lane-level mapping.
Furthermore, Ji Yue’s urban PPA (Point-To-Point Autopilot) function coverage has expanded to a new city, Guangzhou, making it the fifth city to implement PPA after Beijing, Shanghai, Hangzhou, and Shenzhen. In Beijing, where road conditions are more complex, PPA has undergone further improvements, expanding its coverage to encompass major roads within the fifth ring road, effectively covering the city's central area.
