Remote Monitoring and Predictive Maintenance for Industrial Application

The customer is a leading manufacturer of temperature furnaces used in the heat treatment processes for machines components. Process of heat treatment can vary from 4 hour to 24 hour cycle and is based on a thermo-chemical process where the gas is discharged into a sealed furnace with temperature going upto 1000 deg centigrade.

Current Setup:

The machines were connected to the PLC system through sensors to access critical parameters in real time. The system was locally connected to a SCADA system for visualization and reporting. Few parameters are entered manually by the operator.

Requirements :

Customer wanted to explore opportunities with new technologies to be ahead of competition and position themselves as a leading innovator in the industry. The company was looking to enable industry 4.0 principles though the implementation of Industrial Internet of Things (IIoT).

The customer also wanted to remotely monitor the status of the machines, get alerts /notifications for alarm triggers and have a predictive maintenance model to do a future prediction for any anomalies. Customer wanted to monetize this solution by offering it to their customers on a subscription model.

Proposal :

With our in-depth expertise in Industry 4.0, Thingstel proposed services to support customer in making this transition into an Industry 4.0 enabled company. Thingstel proposed a phased Industry 4.0 transformation programme starting with transforming company’s product into an IoT enabled ‘Smart Connected Equipment’. Further through the implementation, Thingstel planned to offer a convergence between the Operational and Information Technologies. Thingstel proposed the use of a partner deice to collect all machine parameters from the PLCs and push it to the cloud for storage, retrieval and further analysis. With the use of a Mobile and a Web Applications operators/users would be able to access important data in near real time. Notifications for alarm triggers would be provided in the application. By collecting data over time, Thingstel plans to build a machine learning model on Google TensorFlow to do process and maintenance related predictions.

Development :

Thingstel successfully completed the development and deployment of the solution on site at one of the locations integrating the PLCs to the backend cloud. Data collected from the machines were stored in data lakes on the cloud for easy access and retrieval.

The solution comprises of ready to use visualisations, reporting, analytics and data science modules categorised into Information Center, Documentation Centre and Production/Maintenance Centre.

The front end was developed on an ionic based web and mobile application for visualization. Alexa integration was also included to read out key information without accessing the dashboard in a secure manner.

Key Benefits :

The solution provided the customer with the following benefits on use as compared to the ones that do not have the solution deployed:

  • Complete Archiving of Process data for Anytime Anywhere access.
  • Predict anomaly and do trend analysis using AI based Machine Learning model to reduce process rejection.
  • Monitor Equipment parameter thresholds to avoid breakdown and thus increase productivity.
  • Access to operation manual and troubleshooting tips on occurrence of specific events tracked.
  • Real time information on running costs, machine idle time and alarm count.
  • Customised platform for the OEM to offer it to end customers on a subscription-based model adding a new revenue vertical to the company of body text.

Conclusion :

The customer is happy with the deployments and the benefits enjoyed and are now looking to start Phase 2 of the development cycle to include more data parameters and provide visualization on the platform provided. Improvements to Information Center, Documentation Centre and Production/Maintenance Centre being taken into consideration.

Return On Investment :

Since the deployment the company has been able to save three complete rejections by spotting the issue earlier. The staff was able to plan better for a heating element failure, which was detected in advance. The team planned for the replacement in advance before the failure could occur and replaced it with minimum downtime.

The historical data available through the Documentation Centre is being used by the Engineering team to better the product and are even considering option to provide access to customer wise information for complete transparency on sub-contract work that they do.

The company is looking to offer the solution as a service on a platform charging customer on a yearly based subscription per asset (furnace). Charging about $1000 per asset per year the customer sees a potential of $10 million business for the 10000 assets they have done the installations for. This will remain a recurring business for the end customers that have subscribed for the service.