[New Wave] Cloud Opens the Era of True Manufacturing 4.0

The Fourth Industrial Revolution, Industry 4.0, and Digital Transformation (DT) have made it increasingly important to integrate IT into traditional industries like manufacturing to enhance agility and create new business models, thereby strengthening industrial competitiveness. This is why cloud computing is gaining attention. It reduces the burden on companies and allows them to focus solely on their core business by using cloud services as needed.

For example, GE Aviation, a jet engine manufacturer, does not just supply engines to aircraft manufacturers but has also launched a cloud-based technical support service for over 450 airlines worldwide. It collects and analyzes data from 5,000 sensors on approximately 15,000 individual engines in operation via the cloud to predict safe operation. Additionally, it has reduced unnecessary maintenance by about 56%, saving $18 million in costs. This is a case of expanding a traditional manufacturing business into a service business using IT.

This is not an easy process for manufacturers. Most factories lack fast internet and infrastructure necessary to immediately integrate IT. This environment is called the edge. Computing capacity capable of creating AI-based machine learning models to store and analyze large volumes of data generated at the edge is essential. However, edge environments often lack such infrastructure, and companies themselves find it difficult to develop machine learning R&D capabilities. This is why cloud-based general-purpose services are attracting attention.

AWS recently launched the Amazon Lookout for Vision service. It is a machine learning service that uses computer vision technology to detect machine defects and anomalies. Factories simply take photos and upload them to the cloud, where the service detects surface quality, color, and shape defects in parts to immediately identify process issues. It is based on high-performance models accumulated from existing parts processes and does not require specialized AI expertise. GE Healthcare, which produces medical diagnostic equipment, applied this service at its factory in Japan.

The Amazon Lookout for Equipment service is also noteworthy. It analyzes sensor data such as vibration, pressure, and temperature from factory equipment to detect abnormal equipment behavior in advance, allowing preventive action before failures occur. Since equipment failure can halt production, early detection of problems is crucial.

Korean manufacturing companies have actively developed and tested these services. For example, GS EPS, a specialist in energy plants, enabled its plant operation team to create AI models without machine learning expertise. Doosan Infracore is also utilizing this for next-generation Internet of Things (IoT)-based equipment production and analysis.

AWS also introduced a device called AWS Monitron, which detects abnormal equipment conditions and predicts maintenance timing. This lowers the entry barrier so that not only large corporations but also small and medium-sized factories can use it immediately. The new year is expected to be the inaugural year for manufacturing-based DT through the cloud.

Yoon Seok-chan, AWS Senior Tech Evangelist