Data analytics: A case for growing analytics maturity
Companies in the process-manufacturing industries have been capturing process data for years, storing it in their historians. But how do you reach the full potential of this data to improve operational performance?
When companies begin their digitalization journey, they go through a series of technical advancements over time. Each of these steps would demonstrate a growing digital and even analytics maturity.
Plants move from a state of capturing data to analyzing it for continuous improvements. As companies begin their digitalization plans for 2022, they can use the following model to go from simply capturing data to a fully functioning smart facility.
Let’s look at what each stage of the model looks like.
The controlled factory
For controlling complex production process, SCADA and DCS systems are being used to signal the control room in case of process upsets and take corrective action. Deeper insight in the process anomalies often is done by downloading the sensor-generated data into Excel and juggling the data with formulas to get insights.
When a historian is added to the mix, some easier insights are gained through the trend client. But for large amounts of data of many sensor readings, these tools lack the ease of use to analyze the data quickly.
The data-driven factory
By putting the power of the data analysis into the hands of the operational experts, companies can make the next step in their digitalization journey. Engineers can analyze the data they know best.
Modern tools use data science capabilities to suggest potential root causes for the process anomalies found. For good behavior, multiple time periods can be overlayed to create fingerprints, which allows the operational experts to monitor operational performance and create soft sensors.
By creating their own analytics-based production dashboards, engineers can make data-driven decisions and a direct contribution to meet business objectives.
The connected factory
When value is gained from analyzing time-series data, a next step in the digitalization journey is made by unlocking data silos. By connecting the data residing in third-party business systems to the time-series data, deeper insights in operational performance can be gained.
A company’s IT department will need to be involved at this level. This stage requires connecting repositories and an analytics platform that allows easy access to all operational contextual data residing in those business applications. Not only is the time-series data put in a broader operational perspective, but also context analytics enables a much deeper insights in production performance.
This creates a wider set of possible use cases and delivers a higher contribution to organizational business objectives, such as reducing carbon footprint, controlling product quality, and increasing uptime.
The augmented or smart factory
Think of an augmented factory as one that has reached transcendence in Abraham Maslow’s Hierarchy of Needs. Augmented factories are the final step in digital transformation.
As data insights and trends become more available, process experts can begin to automate tasks and use machine learning and artificial intelligence to decrease repetition, create anomaly detection models and get prescriptive recommendations to take corrective actions and plan maintenance.
The result? Managing processes only when there is an exception to established norms.
Once experts are managing by exception, they have the ultimate smart factory.
By Edwin van Dijk, vice president of marketing for TrendMiner
