.jpg)
▲Chairman Son Jun-woo of Sonnet is explaining autonomous driving technology.
Comprehensive detection and judgment of objects with 1 lidar and 3 sensors
Deep learning-based algorithm, processing within 10ms, ensuring complete safety
“Safety is the top priority in autonomous driving. The Sonet autonomous vehicle uses radar, lidar, and cameras to detect and comprehensively judge surrounding vehicles, pedestrians, and other objects, and even in an emergency situation, at least one sensor can detect and immediately apply the brakes to prevent accidents. Our technological capabilities are aimed at further perfecting this system, and we will work harder to develop safety technology along with autonomous driving technology in the future.”
Son Jun-woo, Chairman of Sonnet, who met with our reporter who visited Daegu on the 16th to cover autonomous vehicles, cited safety as the top priority in his autonomous driving technology philosophy.
Based on this philosophy, we have been dedicated to developing safe autonomous driving software, and in 2017, we founded Sonnet as the first startup founded by DGIST researchers.
Sonet is a company that upgrades regular vehicles to Level 4 autonomous vehicles and manufactures, sells, and services them in the form of autonomous robotaxis. It is also an artificial intelligence SW specialized company that develops and operates its own full stack of autonomous driving SW.
As if to bear fruit, in November of last year, we obtained a license for paid transportation of autonomous vehicles in Daegu Metropolitan City, and in January, we began providing paid transportation services for autonomous vehicles in the Technopolis area of Daegu Metropolitan City (7.2 km section).
The Sonet autonomous vehicle model that our reporter test drove was the Kona EV, which has one lidar mounted on top of the vehicle and three sensors.
Inside the vehicle, there are installed: △an interface for checking the operation of the autonomous driving system, △a lane recognition camera, and △a signal recognition camera.
There is also an interface behind the passenger seat, so you can check if the system is working properly from the back seat.
The interface allows you to check sensor operation status, vehicle speed, lane and signal recognition camera screens in real time.
The top right of the interface shows whether the sensors are working properly. If there is a problem with any of them, it is designed to switch manually to prevent accidents.
On the left center is a lane recognition camera screen. It identifies whether the lane is a dotted line, solid line, or center line, and a red line is displayed if it is recognized as a lane. It is designed to never cross the center line, ensuring safe driving.
Contingencies on the road are difficult for people to deal with, but the biggest advantage of autonomous vehicles is that they can react to contingencies that people cannot.
The Sonet autonomous vehicle uses radar, lidar, and cameras to detect and comprehensively judge surrounding objects such as vehicles and pedestrians.
Therefore, even in an emergency situation, at least one sensor can detect and immediately apply the brakes to prevent an accident.
The parallel processing capabilities of CPUs and GPUs are very important for quick responses to emergencies.
This ability to process large amounts of data instantly can keep many people safe, including drivers and pedestrians.
For this processing power, Sonnet said it uses industrial PCs with built-in GPUs or embedded systems from NVIDIA.
We have implemented deep learning-based algorithms, such as lane recognition, to process within 10 ms, ensuring maximum safety.
Son Jun-woo, chairman of Sonet, who was in charge of the explanation during this test drive, said that autonomous driving sensors are greatly affected by weather and temperature.
He also noted that optical-based sensors, such as cameras and lidar, do not function properly in situations where visibility is poor due to thick fog, but radar is not significantly affected.
In addition, when driving in a confined space such as a tunnel, the camera is only affected by the brightness, but the radar false detection increases significantly, and the lidar is hardly affected, it was revealed.
In hot summers, the sensors may overheat and not function properly, but Chairman Son Jun-woo said that the influence of the external environment is minimized through technology that comprehensively judges detection results. />
Chairman Son Jun-woo predicted that autonomous driving technology will have a positive impact on future transportation as well as safety.
First of all, he mentioned that the ideal future transportation would be one in which drivers would not lose their jobs but would have an improved working environment in the control room where they can work comfortably, and that autonomous driving technology would transform into a form that provides services.
Based on the autonomous driving philosophy that pursues safety, Sonnet is expected to become a model case for the autonomous driving mobility transport market in Korea in the future by steadily accumulating autonomous driving practical data through technological advancements and paid transportation services.
안전을 추구하는 자율주행 철학을 바탕으로 기술력향상과 유상운송 서비스를 통한 자율주행 실전 데이터를 꾸준히 축적해가는 소네트에 온 국민의 사랑과 무궁한 번영이 함께 하시길 두손 모아 기원합니다^^