Huawei To Make Automotive Sensors For Self-Driving

According to reports, Huawei rotating chairman Xu Zhijun announced at a connected vehicle conference in Beijing today that the company would create an ecosystem of sensors for cars, including radars and LIDAR. Few other details were provided, other than that this would derive from Huawei’s work in 5G technologies and millimeter wave.

Huawei has a history of producing electronics equipment with good quality at lower costs than many competitors. This may bode a future of lower cost sensors for robocars, which would be good. While today, almost all developers are strongly focused on reaching safety goals rather than reducing costs, in the long run all will want to reduce costs. Cheaper sensors can also mean having more of them for greater safety and performance.

Indeed, over time it is expected that the sensor and computing suite in a robocar will cost less than the components removed from a full robocar (like steering wheels, dashboards, pedals, power seats, single panel windshields, mirrors and countless others) meaning that the total cost of a robotaxi is less than the cost of a similar car today. This will bring down the cost of automotive transportation and make it available to a larger population.

Huawei, however, has recently faced battles with the United States, whose officials have expressed distrust of Chinese made technology in critical systems due to the risk of compromise by the Chinese government, a charge which Xu and others have vigorously denied. If the U.S. government or others forbid or pressure against the use of Huawei sensors in cars driven in the USA, this could be problematic. It is less likely a vehicle would be compromised by a radar or LIDAR, but not out of the question.

Radar is an area of strong research today in robocars, because it is low cost and detects things neither cameras or LIDAR can. It is low resolution but improving. While there is no direct connection to 5G efforts in most modern radar efforts, Huawei may simply refer to the value of the expertise they are building up in the high-GHZ bands and extending it to radar. It is also possible they intend to make use of passive radar, where the signal is transmitted from cell towers, bounces off things in the environment and then is received by a vehicle. This is useful technology, but since it only works in certain special places is not a substitute for standard transmit-and-receive style radar. Today’s automotive radars cost several hundred dollars, but there is no reason that models could not be made costing tens of dollars if made in large quantities.

LIDAR is also an area of hot competition and today, high prices. Several Chinese players are attempting to play strongly in a space previously dominated by U.S. companies. Elon Musk has famously been negative on LIDAR, and cost has been one of those reasons. (Indeed, there is no LIDAR available to put in a currently shipping car like the Teslas.) Other players in the space have presumed that low cost LIDAR will arrive and offer superior sensing ability that wins the race to get to sufficient safety levels first.

Disclosure: The author owns stock in Quanergy, a LIDAR vendor, and has in the past advised Huawei senior management on self-driving cars.