Most of the engineers have to rely on existing conventional tools and infrastructure for data conversion and output as the availability of data scientists is limited, says Ruchika Tawani, Data Analytics Engineer, TrendMiner at an E-conference titled, “Empowering Engineers with Advanced Analytics" held on April 29, 2021.

Through the virtual event organized by Indian Chemical News, Tawani explained in detail how democratization of analytics could help accelerate the organizational goals, such as reducing downtime, increasing yield, control quality, improve asset reliability, and increase overall profitability within the process industry. More than an hour-long discussion was moderated by Pravin Prashant, Editor, Indian Chemical News.

As per Tawani, anyone who can understand and use the data, be it process manager, plant managers, process engineers, and IT engineers, could improve operational performance using the TrendMiner software thereby benefiting their respective companies.

She explains, “TrendMiner puts users as a central part of the solution. Once we have two different sets of data, it can do some analysis. There is a dashboard for the user to keep an eye on. There is no single way to find the right path but it tells us how to go about with first sets such as defining the problem. The second is to answer if the situation has happened before and based on that historical data. With TrendMiner, one can also contextualize all the data to address the root cause of a problem. For example, in a pharma batch data analysis under PowderPro 4 week process, the summarized batch information about total batches grade-wise including the rejected ones and golden batches will help in the optimization.”

Sharing an example about batch quality monitoring, Tawani says, “In a polymer production process, different grades of the product are produced (food, industrial, and pharma). During the pharma-grade production, there have been some off-spec batches leading to production losses. For such a problem, our goals shall be to combine lab data and sensor data for a detailed analysis. We need to determine the root cause of the spec off batches besides monitoring future batches for quality deviations. As a solution, TrendMiner uses contextual data to find the low viscosity batches. It shall combine time series data and contextual data. Use a profile of good quality batches as a benchmark for comparison.”

Listing out solutions, Tawani says, “In the Batch Analytics, engineers can analyze good and bad batches to reduce cycle time and improve product quality using golden batch fingerprints. In the case of Continuous Processing, they can embrace the value of digital technology to build operational resilience, reduce cost and increase plant safety. For Asset Analytics, they can make advanced analytics a part of their suite of license to operate tools and significantly improve operational performance.” 

One of the unique features of TrendMiner is that it runs quietly in the background and informs the user in case something is wrong. It runs comparable search results on temperature to detect the root causes for trouble. It offers recommendations on the engine, the correlation between reactor temperature and can create soft sensors, improve asset performance by defining the parameters.

Tawani calls it a versatile software that is user-friendly besides a self-service analytics tool that results in time savings and delivers better quality results than data models. “It is a plug-and-play software with no need for a special setup. It has a user interface with no need for an algorithm to be written by the user but the software itself, based on the data made available to it regularly. Besides robust algorithms for process and asset analytics, it has a highly intuitive user interface.”

Tawani cites the leading pharmaceutical company, Pfizer as one of its biggest customers in the pharma sector. “Trend Miner is specifically developed for process industries. Apart from chemicals, the TrendMinder is relevant to gas and oil, energy, pharmaceuticals, metals, and mining industries,” she concludes while emphasizing the need for quick adoption of the highly productive software.