Mobileye launches EyeQ Kit to support ADAS systems

On July 5, Mobileye, the autonomous vehicle division of Intel Corporation, announced the launch of its first software development kit (SDK) for EyeQ® system-on-chip: EyeQ Kit™, providing global automakers with driver assistance and future autonomous driving technologies. By leveraging the powerful and energy-efficient architecture of the upcoming EyeQ® 6 High and EyeQ® Ultra processors, the EyeQ Kit enables automakers to leverage Mobileye's mature core technologies while deploying their own differentiated code and human-machine interface tools on the EyeQ platform.

Image source: Intel

“With the EyeQ Kit, customers can leverage Mobileye’s proven core technologies and expertise in delivering unique driving experiences and interfaces,” said Professor Amnon Shashua, President and CEO of Mobileye. “As more of the core functionality of the car is defined in software, customers will want the flexibility and ability to differentiate and define their brand through code.”

With EyeQ hardware and software, automakers have access to a broad range of Mobileye solutions, including computer vision, REM™ crowdsourced mapping, and RSS-based driving policies. With the EyeQ Kit, automakers can further leverage the power of the Mobileye system-on-chip to enhance advanced driving features with their brand-unique look and feel. As the visual needs for interaction and communication between drivers and vehicles become more complex, the EyeQ Kit provides automakers with a new way to customize the flow of critical information. The EyeQ Kit will support features such as surround visualization, automatic lane keeping, and road sign recognition through more advanced augmented reality displays.

The EyeQ Kit development team consists of hundreds of Mobileye engineers, bringing together the rich experience of Intel and Mobileye. The development team includes experts in development environments such as compilers and OpenCL standards, making full use of compilation frameworks for intensive computing and deep learning. This approach enables automakers to easily and efficiently develop their own applications. In addition, the EyeQ Kit also opens up the ability to co-host third-party applications, thereby reducing the cost of integrating other chips.

From general-purpose CPU cores to highly compute-intensive accelerators (including deep learning neural networks), EyeQ has a scalable modular architecture designed to achieve high performance while providing the right energy efficiency for deploying AI at the edge in automotive applications. Previous generations of EyeQ chips have been deployed in more than 100 million vehicles, and dozens of global automakers use these chips to provide safety and driver assistance features for hundreds of models around the world.

Designed to reduce development costs, accelerate time to market, and provide hardware vendors with flexibility throughout the development cycle, from concept to deployment and performance tuning, the EyeQ Kit is already being deployed by a major global automaker in a future vehicle program.