18 Jul Workforce 2024: Building Factory Dashboards for Industrial Internet of Things

On Friday, June 28, Workforce Edition continued with a session on Building Factory Dashboards for Industrial Internet of Things (IoT) from Binil Starly, Director of the School of Manufacturing Systems and Network, ASU. The session looked at next-generation dashboards and advancements in data analytics and machine learning.

Factory (digital) dashboards are business or process intelligence communication tools that enable the visualization of factory data from multiple sources. They also convert data into valuable information for those making decisions at all levels within the factory.

Lots of Data

With data in great excess from the growing number of machines with the ability to produce data, it’s important to have a method of bringing the data together in a form that humans can swiftly make decisions on. It should also allow for the running of what-if scenarios to assist with planning out factory/machine line operation with data that’s real time, at least in the last 24 hours.

Long-term predictive and prescriptive analytics are based on the historical collection of data. Because everything is built on data, such analytics can help predict if something is going to fail based on current data collected. Based on actions and interpretations of the humans looking at the dashboards, decision makers can determine what actions can be taken to extend the life or yield.

Dashboards must be customized depending on who is looking at the data and the decision making that needs to result from the data collected. But challenges in the process can quickly arise, and the wrong data can lead to bad decisions. Other challenges that can lead to inaccuracies can include the collection of data in different spaces, the desire to move faster and a lack of assurance as to the accuracy of data collected.

Keeping Data Safe

Starly also cautioned dashboard users must remain cognizant of the security of the data. Insider threats should not be discounted, whether intentional or unintentional. If real-time data is important to decision-making, factories cannot rely on standard cables or ethernet to prevent electronic interference. It’s also important to ensure that data has not been corrupted at any point in the process.

This often necessitates relying on the protocols of other systems. He suggests looking at where data can be mishandled or misrepresented and how to work with providers to lock down the information in a way that is more compartmentalized with backup information. Data organization is also important to ensure elements continue to work correctly.