"Intelligent Driving Technology MAXIEYE" empowers ADAS driver assistance system to achieve mass production and directly hits the pain point of monocular vision ranging accuracy.

At present, autonomous driving is no longer the state of PPT or Demo demonstration. From technological innovation to commercial landing, it is the general trend to build a closed loop of intelligent driving industry and the direction of multi-scene commercial landing.

Founded in 2016, Zhijia Technology MAXIEYE is a leading supplier of autonomous driving products focusing on visual perception technology innovation, and is committed to providing rich intelligent driving products and scene application solutions for different customers and partners in the smart travel industry chain.

On October 20, 2020, the company announced that it had obtained the designated production model project of domestic passenger car Level 2+ intelligent driving system. We will provide customers with Level 2+ intelligent driving System product solutions including high-speed intelligent cruise system based on independent innovation IFVS- intelligent forward visual perception system products and sensor fusion solutions, and open up the technical integration and iteration from Sensing to System based on OTA online upgrade technology, enabling them to upgrade their vehicles intelligently. Previously, the intelligent vision system of MAXIEYE of Zhijia Technology has been mass-produced in the field of ADAS before commercial vehicles.

The illustrations in this article are from MAXIEYE and are used with permission.

Zhou Shengyan, founder and CEO of MAXIEYE, Zhijia Technology, said that the policies and regulations on autonomous driving are gradually improving, which has also increased the policy dividend for the market and promoted it, including the ADAS strong loading regulations for commercial vehicles in 2019, as well as the star rating of Euro NCAP in the global market and various experience requirements for passenger cars, all of which have accelerated the outbreak of the intelligent vehicle market, whether it is ADAS or automatic driving scene landing applications. Driven by the ADAS strong loading regulations of domestic commercial trucks, the ADAS front-loading market of one billion to ten billion levels has broken out.

Up to now, MAXIEYE has launched products and services such as forward intelligent vision sensor system, sensor fusion scheme, ADAS assisted driving system and automatic driving system. In addition, the intelligent driving technology MAXIEYE provides customized product and service combinations according to customers’ needs, such as various intelligent system functions such as ambient brightness detection, rainfall detection, intelligent headlight control and HUD fusion control, and realizes the combination and integration of various functions based on a single device to meet customers’ needs for intelligent upgrading of diversified vehicles.

Since its establishment, Zhijia Technology MAXIEYE has maintained an iterative product development speed every year, and the ADAS system scheme of empowering "eyes+brain" with visual perception has been mass-produced. In 2019, its IFVS-500 series products were released. Based on deep learning, it once again broke through the monocular vision ranging accuracy. Through the integration of virtual lidar algorithm, it can realize the functions of accurate ranging and speed measurement, and support the perceptual prediction of rich target attributes and the construction of three-dimensional scenes. And this year, we will achieve the milestone of shipping tens of thousands of sets in a single month.

The illustrations in this article are from MAXIEYE and are used with permission.

At present, the intelligent vision sensor system of MAXIEYE, a smart driving technology, can realize the detection and recognition of various target attributes, such as lane lines: it can recognize more than seven lane lines such as solid lines, dashed lines and double solid lines, as well as stop lines and zebra crossings, and support three-dimensional output; Driving area detection: it can identify the driving area 100 meters ahead of the vehicle and the slope of the road surface, and accurately predict the types of regional boundaries (such as roadside, vehicles, pedestrians, etc.); Predict the slope and terrain of the road surface, and realize the virtual lane line under extreme conditions such as rainy days and rural roads.

As we all know, there are several conventional sensing sensors in the market, such as monocular vision, binocular vision, millimeter wave radar and lidar. What are the advantages of monocular vision sensors?

First of all, in terms of cost, lidar has not been mass-produced at present because of its high cost, while monocular vision sensor has become the mainstream solution in the mass production market of intelligent driving because of its mature scheme and the best cost performance. Secondly, in terms of perceptual resolution and information richness, vision is obviously better than radar effect. For the function realization of vehicle control system, the vehicle manufacturer pays more attention to the sensing accuracy. Compared with various sensors, both laser radar and millimeter-wave radar belong to active ranging, while monocular vision is criticized for its ranging accuracy, which lacks the technical principle support of binocular vision and one-dimensional depth. However, MAXIEYE, a smart driving technology, can achieve an average accuracy error of less than 5% through its own innovative deep learning algorithm, breaking through the limitation of monocular vision ranging accuracy.

The illustrations in this article are from MAXIEYE and are used with permission.

According to Zhou Shengyan, in order to avoid homogenization competition in the market, Zhijia Technology MAXIEYE maintains an iterative product development speed every year, and with its leading technical capabilities, it has entered the mass production market of car-level pre-assembly. At the same time, through market feedback to optimize products, forming a virtuous circle.

According to reports, the operating income of MAXIEYE of Zhijia Technology has nearly tripled for four consecutive years. Based on the accumulation of previous product technologies, Zhijia Technology MAXIEYE gradually landed in the market from commercial vehicles to passenger cars, and laid out the industrialization application of Level 2+ intelligent driving. At present, the main profit model of MAXIEYE, a smart driving technology, is to deliver forward intelligent vision system with integrated software and hardware and intelligent driving system products based on perceptual fusion scheme for customers.

Up to now, the company’s team size has reached more than 100 people, of which R&D personnel account for more than 60%. Founder Zhou Shengyan studied with Karl Iagnemma, an autopilot scientist at Massachusetts Institute of Technology, and the backbone team of co-founders came from automobile companies and internationally renowned Tier1. He has profound academic accumulation and rich experience in mass production of automobile front-loading projects in the fields of visual perception and autonomous driving.

On the financing level, Zhijia Technology won the Pre-A round of financing from Shengyu Investment in 2019, and announced at the beginning of this year that it has won a new round of strategic investment from Ruili Group, the leading enterprise of commercial vehicle pneumatic braking system in China. At present, A+ round of financing is under way. This round of funds will be mainly used for the technical research and development investment of MAXIEYE for the upgrade and iteration of passenger car intelligent driving system, and provide capital reserves for the large-scale mass production delivery of pre-installed projects.

The headline of this article comes from MAXIEYE and is used with permission. This article is original for the entrepreneurial state and may not be reproduced without authorization, otherwise the entrepreneurial state will reserve the right to pursue legal responsibility against it. If you need to reprint or have any questions, please contact editor@cyzone.cn.