Autonomous driving accidents cause technology to be questioned by software and hardware suppliers

According to the “European Auto News” website, after the Uber self-driving car crash accident occurred, the auto-piloting core technology supplier had a hard time. In the face of all doubts about autopilot sensors and software, these companies can only be exhausted to give explanations. In fact, if the accident did not happen, the above Hardware and software products will soon be brought to market.

Confidence frustrated

Some analysts pointed out that from the above-mentioned corporate response, it can be seen that the systems they are developing are not enough to support more extensive deployment in the market. On the other hand, the failure of similar autopilot tests has caused concern outside the industry that it may drag on the entire industry and, more importantly, consumers and regulators' confidence in the prospects of this technology will also be affected.

Amnon Shashua, Intel’s senior vice president and chief technology officer of Mobileye, a subsidiary of Intel’s R&D autopilot technology, wrote in a blog post about Uber’s auto-pilot car crash: “If there’s more Most similar to the Uber self-driving car accident that killed people, then the already fragile consumer confidence will be further hit, and regulators may be able to kill the future of automatic driving."

The algorithm platform developed by Nvidia has become the basic system of many companies’ automated driving systems, and these companies have suspended their respective similar testing activities within a few days after the Uber self-driving car crash. Nvidia CEO Huang Renxun said last week: “Security is the most important independent topic, and at the same time we are still trying to solve the most difficult algorithm problems.”

Huang Renxun also emphasized that Uber does not use the DrivePX algorithm platform developed by Nvidia for autonomous vehicles.

Uniform standards

Some experts said that autopilot-related hardware such as radar, laser radar, and cameras have such problems, and these problems have exceeded the scope of Uber's own system "drop chain." Manufacturers of the aforementioned hardware cannot comply with regulations if they do not have a set of uniform standards. It is also impossible for regulators to identify how the results of the autopilot test will be safe for future public transport.

The monitoring video content obtained from the Uber self-propelled car crashing into people showed that the sensor set consisting of a camera, radar, laser radar, and laser sensor equipped on the car before the incident could not find pedestrians crossing the road. It is impossible to make a response. The woman who was killed was 49 years old, and she was killed by a Volvo XC90 model that was autopiloted. Some experts have previously criticized Uber's practice of subtracting sensors from Volvo's automatic driving test vehicles currently in use.

However, even before the accident, some engineers warned about the commercial performance of autopilot-related sensors.

JinWooLee, vice president of the Hyundai Motor Group Smart Security and Technology Center, said in an interview with reporters in January this year: “The sensor suite used in self-driving cars is not satisfactory.” He also pointed out existing sensor devices including laser radar. There are limitations on the horizontal and vertical monitoring range.

Kobe Marrongo, chief executive of Israeli sensor startup ArbeRobotics, said in an interview with reporters that sensors currently equipped in autopilot test vehicles have limited many features at the beginning of design, such as adaptive cruising. He said that the current product is not mature enough.

Although various types of equipment on existing test vehicles are used to collect data, improve autopilot performance, and detect obstacles ahead, Kobe-Marungo believes that the above devices should not be fully opened for autopilot functions.

Still in the embryonic supply chain

Some people think that the limitations of the existing autopilot test vehicles highlight the lack of relevant standards. These standards should be formulated by the auto companies and strictly followed by the relevant supplier companies in the R&D process.

Kobe-Marongo told reporters: "All parties are doing their duties, and everyone has their own views on the minimum driving technology."

It is worth mentioning that supply chain companies are trying hard to prove to the public that their new auto-driving technology is to meet the automotive-level safety standards.

Waymo Chief Executive Officer John Kraffirak said at the American Automobile Dealers Association exhibition: "Based on our understanding of what we have seen, we are confident in our own autonomous driving technology, and they are fully capable of handling Uber's automatic driving. The problem that the car faces."

However, for lower-level suppliers who are involved with diversified customers, facing the requirements of different customers on the mileage and performance that the sensors can monitor, how to balance the relationship with the hardware combination will face challenges.

BlackmoreSensorsandAnalytics, a lidar supplier based in Montana, USA, has previously invested in BMW and Toyota. In an interview with the reporter, the company’s chief technology officer, Steven Crouch, said: “The inconsistency of specification parameters has led to a wide range of differences in the results of different systems.”

He said that his company is under great pressure to try to produce advanced sensors at a lower cost and in a shorter time, and this problem is reflected in the fact that the hardware supply chain is still in its infancy.

120mm Push Button

Guangzhou Ruihong Electronic Technology CO.,Ltd , https://www.callegame.com