Follow
Subscribe

Hyundai Develops the World's First Machine Learning-based Cruise Control System

Home > News > Content

【Summary】Hyundai Motor Group announced the development of a machine learning-based Smart Cruise Control (SCC-ML), which the company says incorporates the driver’s patterns into its self-driving behavior, creating a custom experience for the driver. The automaker said its the world’s first AI-based ADAS feature.

Eric Walz    Oct 21, 2019 10:00 AM PT
Hyundai Develops the World's First Machine Learning-based Cruise Control System

Machine learning and deep learning are becoming essential tools in many fields for software engineers, including self-driving cars and advanced driver assist systems (ADAS), which rely on machine learning algorithms to make decisions for safe navigation. South Korean automaker Hyundai has incorporated machine learning in its cruise control system.

The company announced the development of a machine learning-based Smart Cruise Control (SCC-ML), which Hyundai says incorporates the driver's patterns into its self-driving behavior, creating a custom experience for the driver.

The automaker said its the world's first AI-based ADAS feature and it will be available in future Hyundai Motor Group vehicles. The industry-first technology incorporates artificial intelligence (AI) within the vehicle's ADAS.

"The new SCC-ML improves upon the intelligence of the previous ADAS technology to dramatically improve the practicality of semi-autonomous features," said Woongjun Jang, VP at Hyundai Motor Group. "Hyundai Motor Group will continue the development efforts on innovative AI technologies to lead the industry in the field of autonomous driving."

Smart Cruise Control (SCC) is an essential self-driving features that drivers are becoming accustomed to having in today's vehicles. One of the most common features of smart cruise control is to maintain a preset distance from the vehicle ahead while traveling at the desired speed set by the driver.

SCC-ML combines AI and SCC into a system that "learns" the driver's patterns and habits on its own. Using machine learning, the Smart Cruise Control automatically drives the vehicle in an identical pattern as that of the human driver.

Large-38842-HyundaiMotorGroupDevelopsWorldsFirstMachineLearningbasedSmartCruiseControlSCC-MLTechnology.jpg

In current versions of cruise control, the driver manually adjusts his or her driving patterns to the prevailing road conditions, such as selecting a preset distance from the vehicle ahead and choosing a speed. Hyundai's system harnesses machine learning to closely match a human driver's actual behavior. 

The automaker explains that human drivers may accelerate differently in high-speed, mid-speed and low-speed environments depending on road conditions, but detailed fine-tuning of this behavior was not previously available. Hyundai says that without machine learning, it was impossible to fine-tune the settings to accommodate the driver's individual preferences.

This level of fine tuning can make drivers more comfortable using cruise control, thereby improving road safety. With many of today's cruise control systems, some drivers are hesitant to use them since they don't mimic a person's actual driving behavior that well, making drivers uncomfortable at times as they sensed the difference in how the vehicle behaves when it's activated.

Hyundai machine learning-based cruise control uses the vehicle's sensors, including the front camera and radar to acquire driving data and send it to the centralized processor. The computer then extracts relevant details from the gathered information to identify a driver's unique driving patterns. Artificial intelligence is applied during this process to more closely mimic each driver's behavior.

The driving patterns are categorized into three main parts, including the distance from preceding vehicles, how quickly a driver tends to accelerate, as well as responsiveness, or how quickly a driver responds to changing road conditions. The system also factors in current driving conditions and speed.

In a real world situation, a human driver will likely drive with a short distance away from the vehicle in front, while at higher speeds most drivers prefer to be further away from the vehicle in front. This driving pattern information is constantly updated using the vehicle sensors, reflecting the driver's latest driving style. 

Hyundai's SCC-ML factors in these various conditions and makes analysis to distinguish over 10,000 driver patterns, so the technology can adapt to any driver's patterns, as long as they are deemed "safe."

For example, some drivers have a lead foot and may exceed the speed limit or follow too closely to the vehicle ahead. These types of behaviors will not be mimicked. The SCC-ML is programmed specifically to avoid unsafe driving patterns, increasing its reliability and safety, Hyundai says.

When SCC-ML is combined with Hyundai's upcoming "Highway Driving Assist" system that features automatic lane change assist, the automaker claims its new SCC-ML achieves Level 2.5 self-driving.  

Prev                  Next
Writer's other posts
Comments:
    Related Content