The data problem in biomanufacturing

29th Oct 2015

I’ve worked with a number of customers on their vision for a ‘lab of the future’, but what does that really mean for bioprocess manufacturing? Despite the advances in data management technology, the prevalence of paper-based systems in biomanufacturing is still widespread. Quality assurance (QA) and regulatory burdens can make change in these environments slow and labor intensive. But the longer you wait to go electronic – and I don’t mean tracking runs in Excel – the more of your data is lost to the black hole of paper.

At IDBS, we work with customers to capture data with context. It’s vital to allow for traceability across runs and groups, as well as meaningful queries and reporting on the back end. While data management solutions like this have become commonplace in research and are increasingly used in development, the requirements of manufacturing are still spread across a number of systems. This is likely because there is no silver bullet solution to capture all of the requirements for manufacturing – tech transfer, document control, batch records, process control, material and asset management, and holistic data management is no small task.

One could argue that capturing the key process data – run parameters, analysis results, yields and other calculations – would be the most important aspect to manage in a structured electronic system. This is the information that allows for historical data analysis and insight. Every run that is captured in paper or Excel is another that can’t be compared once you implement a solution.

As a process development scientist executing tech transfer to pilot manufacturing, I typically saw QA as a roadblock to change. Every adjusted process parameter or tweaked component required justification before it could be implemented. But while QA can sometimes be a hurdle to change, in this case I’d expect scientists to encourage it. Electronic data management systems stand to positively impact them the most. Imagine how easily investigations could run if everything were in an electronic system with full traceability. Even better – what if the system could prevent some typical deviations from occurring at all, and flag others as they happen so they can be addressed immediately.

No one can say for certain what the lab of the future will look like, but I believe technology – such as intelligent, secure, easily accessible data management systems – will become the foundation for future advancement. You can see evidence today: it’s much easier to migrate data from a legacy Access or other database technology than it is from Excel or paper, which can be impossible aside from manual transcription. These technologies also provide considerable value in the lab of the present, which is just as important.