Sign up for our weekly emails to stay up to date on the latest news!
At the ZF Global Technology Day on June 27, 2024, ZF unveiled its latest innovation, ZF Annotate, a cloud-based AI-enabled validation service for Advanced Driver Assistance Systems (ADAS) and Autonomous Driving (AD) systems. This cutting-edge technology promises to revolutionize the development and quality assurance of these systems by significantly speeding up the sensor data validation process and reducing costs.
Key Highlights:
- ZF Annotate leverages AI to accelerate sensor data validation for ADAS and AD systems by up to 10 times.
- The system is 80% cheaper than manual validation methods.
- Cloud-based service ensures precise, real-time data processing for safe driving maneuvers.
AI-Driven Efficiency and Cost Savings
Modern ADAS and AD systems rely heavily on a multitude of sensors to accurately analyze the vehicle’s environment. Cameras, radar, lidar, and ultrasonic sensors generate a three-dimensional image of the surroundings, which the vehicle uses to perform safe driving maneuvers. The accuracy of these systems hinges on the precise and reliable processing of sensor data, known as the “ground truth.”
ZF Annotate provides this ground truth by using a combination of the customer’s vehicle data and additional ZF sensor data recordings. The AI-powered service processes this data in the cloud, marking, classifying, attributing, and assigning unique IDs to all relevant objects. This automated process forms a highly precise comparative measurement, essential for testing and training ADAS and AD systems.
Dr. Holger Klein, Chairman of the Board of Management at ZF, highlights, “ZF Annotate combines a robust reference sensor set with a scalable cloud service using intelligent 2D and 3D tracking algorithms.“
Accelerated Validation and Scalability
Traditional validation systems relied on manual 2D annotation, which was labor-intensive and time-consuming. ZF Annotate introduces 3D-capable annotation, adding height information to the data, thus optimizing the validation process. This innovation makes the system up to 10 times faster and 80 percent cheaper than previous methods.
Dr. Klein elaborates, “With AI, we can speed up the validation process from 12 months to just two months, saving our customers both time and money.” The cloud-based service operates continuously, ensuring rapid completion of validation tasks without compromising quality.
Flexible Application for Diverse Needs
ZF Annotate offers flexibility in its application. The reference sensors can be mounted on the test vehicle or used in “Pursuit” mode, where a separate reference data vehicle is employed. This adaptability eliminates the need for costly modifications to high-tech test vehicles and allows the service to integrate seamlessly into ongoing development projects.
Klaus Hofmockel, Head of Research and Development in the Driver Assistance Systems and Electronics division, states, “Our cloud-based service completes validation in remarkably short times compared to the market, with no loss of quality.“
Comprehensive and Reliable Validation
The reference data captured by ZF Annotate is not limited to a single viewpoint. Depending on customer requirements, the system can provide a 360-degree view of the vehicle’s surroundings, ensuring a detailed and precise representation. This capability supports the development of ADAS across all vehicle classes, including both passenger cars and commercial vehicles.
Dr. Klein concludes, “ZF Annotate offers reliability and efficiency in a scalable system, paving the way for the software-defined vehicle.” This innovative solution underscores ZF’s commitment to advancing automotive technology and ensuring the highest standards of safety and performance in ADAS and AD systems.
By integrating AI with cloud-based processing, ZF Annotate sets a new benchmark for the automotive industry, promising faster, more cost-effective, and precise validation for advanced driver assistance systems.
Sign up for our weekly emails to stay up to date on the latest news!
