Gerdau and General Electric (GE) have been experimenting with the Internet of Things (IoT) and its extensive application to the manufacturing process. It works by continuously monitoring the allocated assets or equipment, to identify any anomalies from normal operational metrics. After the successful pilot, the functionality has been extended to 600 company assets, from an initial 50 assets. The result is an amazing prevention of unplanned downtime, which yielded cost savings so that no unprecedented additional costs were required. But how did the two companies tap into the potential of IoT and apply it to the shop floor?

Marrying IoT with hard core commercial processes, required a phased approach with a lot of thought put into it at each stage.

Select the right equipment: The first step was to decide which equipment to monitor through IoT sensors. This is critical, as a variety of equipment goes into the steel manufacturing process at varying conditions, such as temperature, pressure and so on. Experts use the method of critical equipment analysis to decide which ones will be responsive to data, at the same time being the core sub-processes for the production process. A lot of study goes into gauging changes in the industrial environment.

Dig Data, Monitoring and Taking decisions: The degree of monitoring also needed to be decided for each of the selected equipment depending on its complexity and on costs. Accordingly, the right data would have to be collated, aggregated and tapped into. It is not necessary that sensors be installed in all selected devices. For example, old devices can be avoided as they can prove costly.

Build a predictive model: The inputs from all these sensors must be built into a predictive model, which forms the overall umbrella for the preventive analytics. It is nothing but a mathematical model, which is devised based on past data.  GE has proclaimed that at least a year’s worth of historical data must serve as the inputs for a worthwhile mathematical model. Herein, the GE’s Industrial Performance and Reliability Center has played a major role.

Translate to action: After the preventive insights are gathered, what do you do with it? Actionable management actions must be devised from them as a step towards continuously improving processes. For this purpose, not only must issue detection be prompt and correct, it must be communicated to the right people at the right time. GE proposes setting up an alert system, which will text any anomalies to the department personnel concerned. These communications can be prioritized and categorized, so that clutter of alerts does not take place.

Gerdau and GE have uncovered phenomenal results from the IoT effort. For example, it was possible to detect the warning signals of a potential breakdown through the mechanism; the company could thus avoid a 72-hour maintenance effort, thereby saving costs in terms of funds and efforts. The baseline for such an application of IoT is extracting, consolidating and analyzing the right data.