Hankook's 'smart factories' advance with AI system

Seoul, South Korea – New artificial intelligence technology developed by the Hankook Tire & Technology Co. Ltd could revolutionise its factory floors and production lines, saving the company from costly shutdowns by predicting a possible problem before it occurs.

Hankook has concluded the research and development of the sensor technology used in CMS+ and is in the process of installing the system in all major equipment in the company's Geumsan plant, located in South Korea.

Using AI and the Internet of Things — the interconnection of the internet and computing devices embedded in everyday objects, enabling them to send and receive data — the tire maker said CMS+ can prevent shutdowns and production line problems through its ability to record a select amount of data, and determine in real time whether the data falls within normal or abnormal conditions.

"We plan to apply this to all of our facilities eventually," said Sungmin Kim, digital technology specialist at Hankook Tire.

That includes the tire maker's plant in Clarksville, Tennessee.

According to Kim, wireless sensors are placed on key pieces of equipment along the production line and record changes in output, vibration, noise and temperature — all in real time. 

And the system does not simply record such data in general terms; the AI technology developed by Hankook is specific to Hankook production lines and machines, meaning it records the vibration, noise and temperature for specific thresholds.

"Various data is collected in real time and used to predict abnormality," Kim said. "CMS+ studies the normal range data and finds those that are outside that 'normal range.' This is possible because Hankook has accumulated 

The AI system also studies fault data, using a separate collection of data from malfunctions, so when similar cases show up, the system will determine that it has an abnormality, Kim said.

"The data collected is not from just one sensor. Different sensors collect data, which are then incorporated together for analysis. The CMS+ system consistently studies all data to create this algorithm," he said.

To secure innovative research and design and digital technology capabilities, Hankook signed an agreement with KAIST, the Korea Advanced Institute of Science and Technology, in April 2019 for joint industry-academic research on future technologies.

As a result, Hankook has made tangible achievements such as developing the Virtual Compound Design system, a predictive model for tire compound properties using AI, as well as the Automatic Inspection Process based on AI and digital sensors. As such, Hankook is implementing tools for its "smart factories" that fuel digital innovation in the overall tire industry.

According to Hankook, the CMS+ system also was developed in conjunction with KAIST, specifically with the university's system intelligence lab.

Professor Jinkyoo Park of KAIST assisted in the development of CMS+.

Park explained that the system uses a three-step AI algorithm, which is analyzed via an IoT module (the sensor), a gateway and a server. With the new system, data analysis is three to four times more precise than the old system, which used sensors with wires and cables and collected huge amounts of data only at regular intervals.

CMS+ and its related algorithm fuse together data from all areas of a production line—IT, equipment, mechanics and timelines for production runs, and uses the information to predict when and where abnormal conditions could occur.

"We needed it to have Hankook's unique expertise, which is why we decided to develop the system ourselves," Kim said. "Such technologies can be developed by an IT or algorithm company, but it will not be perfect."

And with help from KAIST, Hankook said it was able to create a system that best suits Hankook, and no one else.

"It was impossible (previously) to store vast amounts of sensor data transmitted in real time due to limitation in server capacity," Park said. "However, the AI algorithm equipped in the next-generation wireless-based IoT module and gateway made it possible to automatically sort out and selectively store the data suspected of abnormalities."

Kim said CMS+ also is equipped with a real-time alarm system based on wireless communication.

"In the case of anomalies, the system immediately alerts the manager, which allows the manager to take appropriate action much faster," Kim said.

In short, CMS+ predicts the possible abnormality, allowing the problem to be fixed ahead of time and preventing possible machine malfunction. Identifying minor symptoms in real time and performing maintenance in advance play key roles in preventing major facility failures, Hankook said.

"This saves time and energy as the maintenance needed can be executed with minimal loss in time," Kim said. "For example, we ran a test run with a utility that was up for maintenance and the system immediately found an equipment error, which was also confirmed by our experts.

"CMS+ will allow double checking of the system, which will add reassurance in the analysis process, helping our experts make efficient decisions on their reviews."

Hankook said it continues to work with KAIST on the development of further "smart factory" features.