Predictive fault monitoring making Hankook Tire’s plants run more smoothly

In addition to the recently-announced application of artificial intelligence and digital sensor technology to automate the inspection of newly-manufactured tyres, Hankook Tire aims to improve the production flow in its factories via a new monitoring system that utilises artificial intelligence and IoT technology. It calls this Hankook Condition Monitoring System Plus, or CMS+ for short.

Introducing CMS+, Hankook Tire explains that “facility abnormalities” can cause costly and time-consuming shutdowns of entire production lines. These abnormalities manifest themselves in changes to output as well as abnormal increases in temperature, noise and vibration. Identifying minor symptoms that can potentially grow into larger issues in real time and performing maintenance in advance thus plays a key role in preventing major facility failures.

Facility abnormality prediction systems typically rely upon vibration sensors attached to key pieces of equipment. They require experts to analyse the information collected by the sensors and identify the presence of any abnormalities. To improve the accuracy of abnormality detection and to cut down response times, Hankook Tire has turned to artificial intelligence and IoT technology to develop a new facility abnormality prediction system.

CMS+ uses a three-step artificial intelligence algorithm that proceeds through a ‘Next-generation wireless-based IoT module’, ‘Gateway’, and ‘Server’. According to the tyre maker, this arrangement improves the precision of data analysis over the existing system by a factor of three to four.

During the first step that utilises the IoT module, CMS+ collects and analyses data every second, much more frequently than the conventional method. Furthermore, while limitations in server capacities previously ruled out the storage of vast amounts of sensor data transmitted in real time, the artificial intelligence algorithm equipped in the wireless-based IoT module and gateway make it possible to automatically sort and selectively store any data suspected of betraying the presence of abnormalities. This artificial intelligence algorithm was developed in cooperation with KAIST.

At the Gateway and Server levels, CMS+ conducts an in-depth analysis of the data collected based on deep learning technology. It analyses different types of data, including sensor data, temperature, and operational information, to predict in advance abnormal conditions present within the facility. It is also equipped with a wireless, real time alarm system. When an abnormality is detected, the system immediately alerts the plant manager.

Joint research with KAIST

Hankook has already adopted this system within its plants in Korea and is in process of introducing it in other factories around the globe. In addition, it is working to improve the system further by integrating augmented reality in order to more easily identify data flow. In the meantime, Hankook is gradually expanding its adaptation of artificial intelligence through joint research and development with KAIST; the focus here is upon setting up smart factories.

The tyre maker signed an industry-academia joint research agreement on future technologies with KAIST, a renowned science and technology university in Korea, in April 2019. This collaboration has already provided Hankook with tangible benefits, including its ‘Virtual Compound Design (VCD) system’, an artificial intelligence-driven predictive model for tyre compound properties, as well as the aforementioned smart tyre inspection technology.