Outline
Renesas is working on the development of computer vision software and the provision of an automated driving platform that supports functional safety in order to realize a highly automated driving society.
Background
Nowadays, cars and mobility are said to be in the midst of a once-in-a-century revolution, and various functions such as automated driving, electrification, and connectivity are required to be realized. Among them, ADAS (Advanced Driver Assistance System), and its extension, and AD (Automatic Driving) are indispensable functions for realizing a safe and secure society, and automakers are making a concerted effort to develop them. It is becoming more and more common for cars to be equipped with forward-looking cameras and our technology is widely used for image recognition by these cameras. For example, the data acquired by the image sensor is converted to image data by ISP (Image Signal Processor), distortion correction by IMR (Image Renderer), and then recognized as lanes, signs, or pedestrians by image recognition using IMP (Image Processor). As cars become more electrically driven, image recognition is required to achieve not only processing performance but also low power consumption. In addition, automatic driving requires HW and SW that support functional safety, such as stopping the vehicle safely in the event of HW fault or canceling automatic driving and returning control to the driver. Detecting faults while performing image recognition without degrading the processing performance of the system is one of the major challenges in automated driving.
ADAS/AD computer vision system
Concrete Example
Based on the above background, we would like to contribute to the realization of an accident-free society brought about by automated driving by providing an automated driving platform that takes advantage of Renesas's strength in functional safety, in addition to the development of image recognition software using various hardware accelerators.
- Develop computer vision software by using low power consumption dedicated hardware accelerator such as ISP, IMR, and IMP
- Develop computer vision framework for queueing and pipelining of vision processing
- Provide functional safety solution to detect HW fault during autonomous driving
- Provide sensor fusion by combining computer vision and Rader/Lider, and SLAM solution
Sensor fusion for ADAS/AD
Summary
The computer vision software we develop is installed in cars all over the world to support drivers every day. It is very rewarding for us, as drivers, to be able to feel the results of our development close to us, and to contribute to the realization of a safe and secure mobility society through our software. Renesas will continue to develop and provide automated driving platforms to realize a new mobility society and a highly automated driving society.