How Seeq, a Grantek Partner, Predicts Batch Quality at Life Sciences Manufacturing Facilities
March 21, 2022
Grantek’s partner Seeq is proud to share the below use case. Grantek became a Seeq partner in 2019. The partnership helps Grantek better serve our clients by accelerating digital transformation and harnessing diagnostic and predictive analytics to enable greater operational efficiency and deeper insight across their facilities and enterprises.
Founded in 2013, Seeq publishes software applications for manufacturing organizations to rapidly find and share data insights. Oil & gas, pharmaceutical, specialty chemical, utility, renewable energy and numerous other vertical industries rely on Seeq to improve production outcomes, including yield, margins, quality, and safety. Headquartered in Seattle, Seeq is a privately held virtual company with employees across the United States and sales representation in Asia, Canada, Europe, and South America. For more information, visit seeq.com.
Nothing is more important than protecting patient health. That is why quality is the most critical metric in pharmaceutical manufacturing. During manufacturing of new or existing medicines, drug companies need to test each batch to ensure that the quality consistently meets standards.
Predicting the quality of each batch is a challenge for most drug manufacturers. It is a labor-intensive and time-consuming—though necessary—process. Typically, samples are taken and sent to the lab for analysis while the process is actively running. The analysis alone adds several hours to the process time. And, if the lab returns inadequate results, time-consuming—and often expensive—changes need to be made if the batch is recoverable. If not, the manufacturer can lose hundred of thousands to millions for the lost batch.
Real-time predictions are the goal. A large molecule pharmaceutical company was struggling to make near real-time predictions. Unfortunately, delayed lab results made it quite difficult for them to optimize the process inputs to control the batch yield. By setting process inputs without optimizing the process, energy and raw materials can be wasted, and product quality and yield may be reduced. The company needed a more efficient way to predict batch quality for process optimization.
Using Seeq, the scientists running the processes built a model of process quality based on data from the OSIsoft PI data historian. The manufacturing team uses this model to predict the quality of the in-progress batches. This allows for modifications to be made during the production process before the batch would be lost due to quality issues.
The analysis used takes standard process measurements such as the reactor temperature, volume, and concentration as process parameters for controlling yield. The raw data is filtered to suit the desired operation of interest: the reactor heating portion of the process. Using statistically significant process parameters, a predictive model for yield is generated. The model was then deployed to detect abnormal batches during their manufacturing processes.
The manufacturer has potentially saved millions of dollars through process optimization. They are no longer waiting for quality tests to return from the lab before making decisions. Instead, they are able to rapidly identify and analyze root cause analysis of abnormal batches via modeling. Thus reducing the number of out-of-specification batches by adjusting process parameters during the batch and lowering the amount of wasted energy and materials.
Developing and deploying an online predictive model of the product quality and yield can aid in fault detection and enable rapid root cause analysis, helping to ensure quality standards are maintained with every batch.