Modern automated driver assistance systems (ADAS) require a plethora of sensors to accurately analyze a vehicle’s environment and ensure safe driving maneuvers. To enhance the development of these ADAS and autonomous driving (AD) solutions, technology giant ZF has introduced a cloud-based, AI-enabled validation service named ZF Annotate.

ZF Annotate Leverages AI for ADASThis innovative service was unveiled at the ZF Global Technology Day on June 27, 2024, in Hanover. This event marked the debut of a fast, cost-effective, and market-ready service tailored for both passenger and commercial vehicles.

The Crucial Role of Accurate Data in ADAS Development

Accurate and reliable data are indispensable for the development of advanced assistance systems in vehicles. A myriad of sensors, including cameras, radar, lidar, and ultrasonic sensors, continuously supply information, enabling the vehicle to create a three-dimensional image of its surroundings. These systems must identify a variety of objects in real-time, such as vehicles, pedestrians, lanes, and traffic signs.

For the vehicle to perform correctly, this sensor data must be digitally processed to provide what is known in the industry as “ground truth” – an accurate depiction of reality. Comparing the collected sensor data with a reliable, high-precision reference sensor set enhances the accuracy of this ground truth, which is where ZF Annotate plays a pivotal role.

AI: Enhancing Speed and Safety

ZF Annotate utilizes both the customer’s vehicle data and additional ZF sensor data recordings to provide a “ground truth.” This cloud-based service acts as an independent redundant setup, receiving the same information as the vehicle on the road. The recorded data is then uploaded to the cloud for analysis.

Using artificial intelligence, the system accurately marks, classifies, attributes, and assigns unique ID numbers to all relevant objects, tracking moving objects as well. This object information contributes to a comprehensive environmental model, forming the ground truth.

Following this annotation, the software provides a highly precise comparative measurement, making ZF Annotate a cutting-edge, AI-supported validation solution for testing and training modern ADAS/AD tech from Level 2+ to Level 5.

“ZF Annotate combines the robustness of an independent reference sensor set with the scalability of a cloud service, employing intelligent 2D and 3D tracking algorithms,” states Dr. Holger Klein, Chairman of the ZF Board of Management.

A Unique, Intelligent ‘Cloud Factory’

Previous systems primarily relied on 2D annotation for reference data validation, mapping the environment in terms of distance and horizontal angle. In contrast, the 3D-capable ZF Annotate adds height information, making it a unique and intelligent ‘cloud factory’ in the industry. This advancement optimizes a previously labor-intensive and costly service in multiple ways, as Klein elaborates.

ZF Annotate Leverages AI for ADAS

Accelerating Development: Up to 10 Times Faster and 80% Cheaper

The precise and reliable data from ZF Annotate’s reference measurements can significantly expedite the development and fine-tuning of complex ADAS and AD systems. Traditionally, validating these systems was a laborious, time-consuming, and costly process, as the reference data was manually annotated.

“Artificial intelligence allows us to speed up the validation process by up to ten times, reducing the annotation period from 12 to just two months, saving our customers both time and money,” Klein explains. Klaus Hofmockel, Head of Research and Development in the Driver Assistance Systems and Electronics division, adds, “With ZF Annotate, we can generate a ‘ground truth’ rapidly, working 24/7 to complete the validation of reference data swiftly and without compromising quality.”

Flexibility and Scalability: Adaptable to Customer Needs

ZF Annotate’s reference sensors can be employed directly on the test vehicle or in “Pursuit” mode, where the sensors are mounted on a separate reference data vehicle. This flexibility allows for the evaluation of various scenarios, such as parking in public spaces, without significant adjustments to the vehicle under test.

This versatility makes ZF Annotate independent of specific sensor manufacturers, eliminating the need for costly modifications to customers’ high-tech test vehicles. This capability enables customers to incorporate the service into ongoing development projects, supporting the validation process and helping developers achieve flawless products.

Moreover, the recorded reference data is not limited to a front view. Depending on customer requirements, the reference sensor set can provide a comprehensive 360-degree view, offering a detailed and precise representation of the vehicle’s surroundings.

A Revolutionary Step Toward Software-Defined Vehicles

ZF has a history of developing cutting-edge technologies across various application areas. ZF Annotate is applicable across all vehicle classes, including both passenger cars and commercial vehicles.

“Reliability and efficiency are crucial for validating reference data in the development of advanced driver assistance systems,” Klein notes. “ZF Annotate offers both in a scalable system, representing another milestone on the road to the software-defined vehicle.”

Conclusion

ZF Annotate represents a significant advancement in the development and validation of ADAS and AD technologies. By leveraging artificial intelligence and cloud-based services, it provides a faster, more cost-effective solution for creating the ground truth necessary for these advanced technologies. This innovation not only accelerates the development process but also ensures a higher level of safety and reliability, paving the way for the future of autonomous driving.




Tags: , , , , ,

Related Article

No Related Article