Google Cloud Japan and WingArc1st Collaborate to Support the Digital Twin of the Manufacturing Industry

WingArc1st and Google Cloud Japan will start offering a solution for the manufacturing industry in Japan today that integrates WingArc's BI dashboard MotionBoard and Google Cloud's Manufacturing Data Engine.

Background
The environment surrounding the Japanese manufacturing industry is changing on a daily basis, with chronic labor shortages, diversifying consumer needs and the consequent shortening of product development cycles, as well as the need to restructure the supply chain as a result of COVID-19. In order to predict and respond quickly to such changes, DX utilizing data from inside and outside the company has become an urgent need for both management and the field. In particular, the concept of the digital twin, which is a virtual digital replication with 3D maps, etc., used mainly in the manufacturing industry by major companies, is an important management strategy from the viewpoint of predictive simulation and cost benefits through data visualization. However, in order to collect and visualize the necessary data,  it is necessary to solve issues such as skill shortages amongst engineers and the replacement of aging equipment.

Summary of the integration
Google Cloud's Manufacturing Data Engine serves as the IoT data foundation for the digital twin, normalizing and hierarchically organizing data generated on the production floor. MotionBoard, which specializes in 3D modeling and camera integration, visualizes the data.

An edge solution that collects data using more than 250 industrial communication protocols from Japan and overseas is also provided, and data collection from various production facilities can be started immediately with only a few settings. Scalable system construction is possible by utilizing the cloud, allowing for economic system investment by starting with a small-scale system and checking the effects of its introduction.

Data can be collected not only from IoT devices, but also from enterprise-level systems such as ERP and production management systems, and this data can be aggregated into the Google Cloud analysis platform. By correlating data from production sites with KPIs for factory management and visualizing them via MotionBoard, the system supports prompt on-site actions and management decisions based on the data. It is expected to contribute to the promotion of DX in the manufacturing industry by facilitating data utilization by on-site users and accelerating the democratization of data.

Major usage scenarios

・Visualization of carbon footprint and the carbon neutrality
We measure energy usage for each process and measure energy usage per production instruction to calculate the carbon footprint of each product (single unit or lot). By making design changes and process parameter changes to consume less energy, the goal is to optimize energy usage without affecting quality while maintaining productivity. It also uses AI to optimize control loops, which is expected to reduce energy consumption and maximize yield.

・Centralized management and analysis of factory data
Not only IoT data, equipment sensors, production data, and maintenance history from manufacturing sites, but also various data from sales and management departments, such as financial records and case information, can be aggregated into the Google Cloud analysis platform. By integrating and visualizing OT data and management data, the isolated islands of information that arise between fragmented organizations can be eliminated, enabling quick, data-driven management decisions.

・Quality inspection and analysis of non-conformance factors by AI
The system automates visual inspections utilizing cameras and AI, displaying the results on the screen along with the captured images. Since the inspection results are converted into data, the relationship with data related to upstream processes can be analyzed to identify non-conforming factors, enabling the automation of various factory operations, such as streamlining quality assurance processes and optimizing production, using AI and machine learning technologies.

 ・Process Anomaly Detection and OEE Predictive Maintenance
Applying AI for anomaly detection, the system detects anomalies in time-series data streamed on the data infrastructure with only a few settings. The results are displayed on a dashboard, prompting operators to take countermeasures. In addition, models can be created to diagnose signs of poor production performance and downtime, and economical maintenance can be performed at the optimum time to maximize OEE and potentially eliminate unexpected equipment stoppages.

 ・Integration of factory data to achieve mass customization
The integration of the engineering and supply chain will constitute a digital thread. It is expected to realize product offerings quickly and economically through mass customization, responding to customer requirements in detail. By sharing the digital twin data provided by the database, progress will be visually monitored in detail from order to delivery.

Online seminar