The development of technology requires money, and investors must also see tangible results.
Creating an L3.5 autopilot that suits China's complex road conditions has always been something Ni Kai wants to do, but commercializing technology-enabled products can be a realistic basis for realizing the dream of autopilot.
Just announced the completion of tens of millions of dollars in financing Wo Technology recently released its first generation of autonomous driving R & D platform, it has a very Chinese name: Regulus.
'We were in the autopilot R & D trip a lot of pit, Xuanyuan platform is to make the present autonomous driving R & D units less detours, Xuanyuan efforts to empower the limbs (body) and eyes (sensor) of the development platform.We hope that through the Xuanyuan And colleagues to share a pro-test effective solution, we used to fill the pit of precious time and energy liberated.'Nikai said that customers only need to Xuanyuan platform for autonomous vehicles to the brain, which is the so-called decision-making system.
Autopilot platform, we may think of the first time this year, released the Apollo Baidu high profile platform, but Ni Kai stressed that the Regulus and Apollo is not the same, Regulus also want to create that is just mentioned autopilot limbs and Eyes, customers take the past need to re-develop, are semi-finished products, while Apollo perception from the decision-making is a complete set of self-driving solution, that is to use.At the same time, the two sides coexist, in some ways is to reach a cooperation .
Zhi Jiajun understands that the Xuanyuan platform supports level L2-L5 auto-driving from two levels of software and hardware. The platform hopes to give the basic capability of automatic production of the traditional mass production cars and provide customers with power supply, cooling design, sensor coordination, and control stability These are time-consuming and systematic projects.
In the product manual, we see that Xuanyuan can provide software and hardware solutions such as body longitudinal control, lateral control, wire control and hardware interface, etc. Xuan Yuan will open up the three capabilities of sensor / computing equipment, basic software and remote control.
Wo Wo also gave us a list of supported hardware and software of the Xuan Yuan platform, including laser radar, millimeter wave radar, CAN gateway, middleware and so on, as can be seen in detail:
Currently, the Regulus platform has supported the public MAGOTAN, Buick Angkorway, the future can be customized according to user needs services to support more models of wire-based.
Outside the press conference, we saw the two modified cars, the first impression of the presence of the two cars was: clean, clean, not the same as our usual self-driving car, full of radar With camera-aware hardware, the look of the two cars is almost the same as the one we see on the road everyday, with no non-heterogeneous appearance.
There is a camera in the front of Enkewei, an antenna is attached to the roof shark fin, and other sensors are not known. The whole body of the MAGOTAN has six laser radars in addition to the same camera and antenna. However, compared with the appearance of the traditional autopilot, Image is not unexpected.
Unlike many self-driving companies that hid themselves and staff open the magotan trunk generously, we were not able to sniff out too much 'secrets' in the trunk as children's bicycles ran through them.
Sharp-eyed people will find that this Maoteng trunk still has some clues in the trunk cover a heat outlet exposing the core, the entire unit is likely to hide in the spare tire position, the car from the outside is very Clean, a folding bicycle can easily be put.
Of course Wo Wo staff did not intend to open the cover, only revealed that the initial integrated hardware design has been completed. Unfortunately, we did not see the true nature of the entire system can only look forward to the next Wo more burst more violent However, compared with many self-driving cars messy trunk, so neat and rare, whether it is appearance or internal, Wo multi-technology team seems to have some obsessive-compulsive disorder, engineering capabilities can not be underestimated.
Wo more than two cars set up three links. 8 word winding pile - Angkowei, not brake emergency lane change - MAGOTAN, no one connected - MAGOTAN.
8-word wound pile difficult lies in the precise control of the vehicle, and body control is the Xuanyuan platform to solve the pain points for customers.
The entire demonstration process is not long, because it is a test car, Angkwei's control switch has not yet been implanted in the car, the staff need to iPad in the hands of the iPad click the Start button to start the remote control, then the AD system to take over the vehicle, driving Speed is very fast. And human pilots, the system also intends to corner 'throttle', a small corner speed reached 30 yards, Chi Jiajun in the back even a bit 'motion sickness', the entire corner Action comparable to the old driver, quite suspicious of dazzling.
However, it should be noted that the demonstration of this 8-bend is only based on the established route of positioning. Therefore, Sensor is not a part of the show here. The most important manifestation is the skill of the Xuanyuan platform for body control.
Not brake emergency lane change will test the track above, by the MAGOTAN protagonist. Emergency lane change - do not brake - then re-line back to the original lane, the system needs to complete this action instantaneously, the entire journey speed reached 55km / h. Steering and throttle coordination is very tacit understanding, ordinary drivers in the usual car are very difficult to accomplish such an operation.
Such an emergency scene as a simulation of China's complex road conditions is a simulation of multi-technology research and development in the Xuanyuan platform 'localization' performance.
In the last one of the unmanned connection is relatively less show off, the vehicle loop circuit, moderate speed, the whole more than a kilometer, a total of more than a dozen sites.The test is the vehicle starting, parking and driving Stationarity, this scenario can be seen as the operation of unmanned shared car simulation, the application prospects are very large.
With a clever brain that also needs strong limbs and bright eyes, vehicle control plays a decisive role in autopiloting the entire system, and as a start-up, there is a need for a self-hematopoietic ability in addition to a predictable future that can be built using existing technologies Sales of basic products to customers in need is a wise move, Xuanyuan platform can be regarded as such an outcome.