Kia Motors, South Korea's oldest automaker, has announced that it will develop adaptive cruise control, the world's first Smart Cruise Control Machine Learning, together with other members of the Hyundai Motor Group (HMG). It's about technology, which incorporates the driver's samples of the autonomous driving mode to accommodate the driver.
Technology, the first of its kind in the automotive world, as part of the Driver Assistance Assistant (ADAS) system will include artificial intelligence (AI). The system will be gradually integrated into future Kia vehicles, as well as the Hyundai and Genesis sisters, and will be given the Genesis GV80 as the first model. Adaptive cruise control is a prerequisite for autonomous driving and is essential for ADAS as it maintains distance to the vehicle in front while driving at the speed set by the driver.
The SCC-ML combines AI and SCC into a system that learns from the driver's driving habits. Machine learning adaptive cruise control drives autonomously with identical patterns as the driver.
With today's cruise control, the driver manually adjusts the driving patterns, such as eg. distance to the vehicle in front and acceleration. Therefore, it is very difficult today to fine tune the SCC according to the wishes of the driver without using machine learning technology.
For example, the same driver accelerates differently at high, medium, and low speeds due to different circumstances, but no more precise tuning is possible. Therefore, when the radar cruise control is activated, the vehicle responds differently than desired, which drivers feel and because of the inconvenience of avoiding the widespread use of this technology.
The group itself developed SCC-ML, which works as follows: first the sensors (eg front camera and radar) constantly collect driving information and send it to a central computer. It then extracts important details from the information collected to identify driver patterns. During this process, artificial intelligence technology called machine learning algorithm is used.
The driving pattern consists of three parts: distance to the vehicle in front, acceleration (how fast the vehicle accelerates) and response (how quickly it responds to driving conditions). In addition, driving and speed conditions are taken into account.
It collects data when keeping a short distance to the vehicle ahead during slow driving, city driving and while accelerating on the lane. Considering these situations, SCC-ML analyzes more than 10,000 samples and thus develops flexible radar cruise control technology that adapts to driver habits.
Driver sample information is regularly updated with sensors that respond to the latest driver style. In addition, the SCC-ML has been programmed to not take into account dangerous driving habits, which is why it enhances its reliability and safety.
For the time being, the SCC-ML system does not allow tuning for two drivers in one vehicle, but after only one hour, the system adapts to the other driver. In the future, the group plans to extend the system to curvy roads or when changing lanes. With the incoming Highway Driving Assist system, which includes the automatic replacement of the SCC-ML lane, it will start at a level of 2.5 autonomous driving.