LAS VEGAS--(BUSINESS WIRE)--MINIEYE, a Chinese company focusing on sensing technology of autonomous driving, today announced a technology collaboration with Xilinx, Inc. the leader in adaptable, intelligent computing, to develop turnkey sensing systems for advanced driver-assistance systems (ADAS). The two companies are working together to create solutions for carmakers and Tier-1s by running MINIEYE IP on the Xilinx Zynq-7000 and Zynq UltraScale+ auto-grade MPSoC platforms to enable sensing features for L1-L3 autonomous driving.
According to Liu Guoqing, founder and CEO of MINIEYE, the new technology is able to recognize and analyze more than 20 traffic objects using Xilinx’s auto-grade MPSoC platforms. The solutions enable in-cabin detection, object detection, as well as the signal input of other sensors. These functions are designed to adapt to complicated situations inherent in autonomous driving. Leveraging the technological advantages of MINIEYE’s algorithm and Xilinx’s advanced platforms. MINIEYE has already installed its before-market product into many vehicle types of mainstream OEMs.
“MINIEYE has already proven its ability to create high-precision, highly efficient AI and deep neural networks for automotive on Xilinx technology,” says Willard Tu, senior director of marketing, automotive, Xilinx. “In our collaboration with MINIEYE, we hope to leverage its ability to offer turnkey solutions for a variety of automotive applications such as mono front camera sensing and interior cabin sensing.”
“Currently, Mobileye is by no doubt the top one in the market. It seems that OEMs and Tier 1 suppliers have little choice. If not choosing the chip developed by Mobileye, OEMs and Tier 1s have to invest huge resources in algorithm development. The cooperation between MINIEYE and Xilinx aims to bring more market choices for OEMs and Tier 1 for vehicle-mounted sensing solutions.” Liu said.
The solution enables customers to deploy differentiated visual algorithms and custom algorithms integration from MINIEYE or others on the same Xilinx-based technology. What’s more, Xilinx SoCs are both hardware and software programmable, enabling new features and functions to be added after the vehicle has been deployed to meet NACP’s updating requirements in different countries.
Dr. Liu also reveals the timeline. The IP for L1 and L2 has been applied in several vehicle types and will achieve mass production within the year while the L3-oriented counterpart will be launched in late 2019.