The technologies, first of its kind, features artificial intelligence (AI) in the Advanced Driver Assistance System (ADAS) features. The system is planned for implementation in the future of Hyundai Motor Group automobiles.
“The new SCC-ML improves with the intelligence of previous ADAS technology to drastically enhance the practicality of semi-autonomous features,” stated Woongjun Jang, VP at Hyundai Motor Group. “Hyundai Motor Group is going to continue the development projects on revolutionary AI technologies to lead the industry in the area of autonomous driving.”
Smart Cruise Control (SCC) allows an important self-driving feature and core technology for ADAS: keeping a distance from the vehicle forward while going at the speed selected by the driver. SCC-ML brings together AI and SCC into a method that learns the driver ‘s habits and patterns on its own. Through machine learning, Smart Cruise Control autonomously drives wearing the same pattern as of the driver.
To be able to run the previous Smart Cruise Control, the driver manually adjusted driving patterns, like the distance from the preceding vehicle and acceleration. It was extremely hard to meticulously fine-tune the adjustments to support the diver’s specific preferences with no machine learning technology.
For example, including the same driver might accelerate differently in high speed, low-speed and mid-speed environments based on condition, but the detailed fine-tuning wasn’t available. Thus, when Smart Cruise Control was triggered and the automobile operated differently than they prefer, drivers, sensed the difference, leading to a reluctance to make use of the technology since it made them feel unstable and anxious.
Hyundai Motor Group’s independently produced SCC-ML works as follows: First, sensors, like the front camera and radar, continuously acquire driving info and send it to the centralized computer. The computer then extracts related details from the gathered data to determine the driver ‘s patterns. An artificial intelligence technology known as machine learning algorithm is used during this process.
The driving pattern could be categorized into 3 parts: distance from preceding automobiles, acceleration (how fast it accelerates), and responsiveness (how fast it does respond to generating conditions). Additionally, driving conditions plus speeds are considered as well.
For example, maintaining a brief distance through the preceding vehicle throughout slow, city driving, plus even further away when traveling in the fast lane. Looking at these different problems, SCC-ML makes analysis to distinguish more than 10 thousand patterns, creating an adaptable Smart Cruise Control technological innovation that could adjust to any kind of driver ‘s patterns.
The driving pattern info is constantly updated with sensors, reflecting the driver’s latest driving style. Additionally, SCC-ML is programmed especially to stay away from learning unsafe driving patterns, boosting its safety and reliability.
With the upcoming Highway Driving Assist method which features instant lane change assist, SCC-ML achieves Level 2.5 self-driving.