Current location - Loan Platform Complete Network - Big data management - Accelerating the layout of automatic driving, Toyota led the investment in two Chinese companies, released the automatic driving simulation platform
Accelerating the layout of automatic driving, Toyota led the investment in two Chinese companies, released the automatic driving simulation platform

In the past, Toyota has always been a conservative brand, embracing technology is more like Audi, BMW and other German car companies are willing to do. However, Toyota's latest move in the field of automated driving has been frequent, not only in the domestic investment in two automated driving startups Pony Smart and Momenta, and cooperation with the Massachusetts Institute of Technology Laboratory of Computer Science and Artificial Intelligence has also made progress again with the release of a new automated driving simulation platform "VISTA".

Momenta to provide high-precision maps for Toyota in China

Toyota on March 18 led an investment in automated driving startup Momenta, which will provide high-precision maps and update services for Toyota's camera vision technology and advance the commercialization of Toyota's Automated Mapping Platform (AMP) in China.

High-precision maps will be a key part of realizing L3-level automated driving. Before L3-level automated driving, monitoring the environment is mainly done by the driver, while L3-level automated driving is started by the system to replace the driver's overall environment monitoring, evaluation, decision-making, and realizing conditional automated driving, and the key to the evaluation is the need to use high-precision maps. For example, high-precision maps allow the vehicle to obtain lane-level positioning and information that can be projected and supplemented for the 1-kilometer road ahead, and can play a major role in curved roads and rainy and foggy weather in particular, providing full redundancy for system safety.

Momenta's high-precision maps are based on a vision solution that combines deep-learning perceptual technology with vision-based map building and localization technology (SLAM), which enables automated, high-precision map building with a relative accuracy of up to 10 centimeters using only consumer-grade sensors (cameras, GPS, and IMUs), meeting the accuracy requirements for automated driving at Level L3 and above. At the same time, Momenta's automated map production line can also efficiently process big data to ensure the accuracy of map information.

Leads investment in Pony Smart, starts self-driving pilot project

On February 26, Toyota led the investment as a strategic investor, again contributing $400 million (about 2.835 billion yuan),? expanding the cooperation between the two companies in the field of mobility. In fact, before this investment in Ponywise, Toyota has established cooperation with Ponywise in August last year, announcing *** with the start of self-driving pilot projects on open roads in China.

It is understood that Toyota's investment this time is an extension of the basis of the last business cooperation, this time will be in addition to the previous agreement *** with the development of automated driving technology, will further accelerate the research and development of automated driving and commercialization applications, to explore more possibilities in the field of travel.

Pony.ai is a company that specializes in autonomous driving solutions, and its product Pony.ai is based on radar, optical radar, GPS and computer vision technologies to sense its environment for the purpose of autonomous driving. Pony.ai is also the first company to provide self-driving mobile cabs (Robotaxi) in the Chinese market, releasing the PonyPilot self-driving travel service project in December 2018, taking the lead in piloting urban open roads in Nansha, Guangzhou, to start regular operation, and in November 2019, it took the license for commercial trial operation in California, USA.

PonyPilot has accumulated more than 70,000 orders by December 2019, and the trial operation during this period has undoubtedly provided a complete test of all aspects of self-driving travel, including self-driving technology, vehicle scheduling, finding a parking space, parking, etc., and it also allows for the evaluation of all aspects of performance through user feedback. It can be said that Ponywise's self-driving travel service is an effective way to verify the maturity of the technology, and Toyota is being able to accelerate the implementation of self-driving technology through the cooperation of Ponywise.

Autonomous driving simulation platform "VISTA", data from real-world environments

Additionally, Toyota has made the latest progress in its collaboration with the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory (MIT?CSAIL), developing a new autonomous driving simulation platform, "VISTA". "VISTA". The platform uses datasets of cars driving in the real world, eliminating the need for humans to manually mark road signs, lane lines, physical buildings and other information, dramatically optimizing self-driving car testing and deployment times.

During the development of self-driving technology, road testing is the most time-consuming and labor-intensive process, although it is less difficult, and it is undoubtedly a shortcut to move testing to a computer to "speed up" the process. Tesla CEO Musk once said that at least 6 billion miles (about 9.65 billion kilometers) should be tested before self-driving cars are safely put on the road. If you want to complete these tests for real, it would take an inestimable amount of time.

In fact, simulation platforms have always been an integral part of the automotive development process, allowing components or cars to "speed up" in simulated environments built in computers, so that defects can be quickly identified and adjusted, and then supplemented by road tests. Moreover, in order to ensure that the simulation environment is the same as the real world, the data used by VISTA is extracted and analyzed frame by frame from the video data collected during the actual driving, which is consistent with the real road conditions and the targets involved in the driving situation.

Conclusion

Prior to this, Toyota's actions in the field of automated driving, although not much, but its venture capital company AI?Ventures, has actually quietly invested in nearly 20 automated driving-related startups, covering a wide range of related fields. However, Toyota in the last year, a change in the previous low-key style, publicly launched the Lexus LS?500h based on the creation of the self-driving test car, and recently invested in a succession of pony Zhixing Momenta two Chinese companies, and released the automatic driving simulation platform "VISTA", is clearly the automatic driving thing to the surface to accelerate the field of automatic driving. The Chinese market is likely to become its main launching point.

This article comes from the author of AutoNavi, and does not represent the views of AutoNavi.