Driven by scientific and technical innovations, an explosion in available data, and the need to get faster and more effective, the migration of biopharmaceutical manufacturing to a digitally mature environment is a high priority. Standards and best practices are emerging, but few organizations effectively leverage these tools. At their core is the data required to drive improvement.
Key challenges are that data is scattered across multiple systems, it exists in various formats, and it is difficult to assemble in the proper context for evaluation, reporting, and decision support. Also, personnel spend significant amounts of time gathering and contextualizing data rather than using it.
A better approach to data management is needed to upgrade biotech and pharma operations to the conceptual standards of pharma 4.0, building on connected plants and instruments to enable real-time process monitoring, quality control, and product release.
A critical first step toward better data management is establishing a standard ontology for biomanufacturing. We have therefore published Big data to smart data: implementing an ontology and digital data capture to improve biomanufacturing. This defines the critical concepts for describing and capturing a manufacturing process independent of the systems and equipment used.
We reviewed best practices for ontology development across industries and adapted them to the biomanufacturing space. Then we carried out an initial proof-of-concept implementation of ontology-enabled analysis at the North Carolina State Golden LEAF Biomanufacturing Training and Education Center (BTEC).
As part of the proof-of-concept, BTEC underwent elements of a digital transformation to enable better data measurement and capture to ensure the ontology implementation would be successful. On this digital foundation, BTEC implemented BioPhorum’s standardized ontology that enabled the synchronization of online and lab data measurements to calibrate optical density probes across five bioreactors.
Using this ontology-based software reduced the duration of the calibration workflow from 4 hours to 30 minutes.
The proof-of-concept work demonstrated how technologies can help organize data and retrace its provenance to source systems and time of acquisition. This is a fundamental prerequisite for quality control and can be extrapolated to illustrate the enormous potential in real-world scenarios where each unit operation can use several new in-line sensors for real-time quality control.
The results at BTEC demonstrated how implementing better data management as part of a digital transformation can create compounding value over time. Our paper demonstrated that leveraging an ontology, combined with the digitalization of a bioreactor setup, can unlock and accelerate improved analytics, operations, and production. It is an important step toward better data management. For more information, download the paper here and contact Tim Horton, Senior Global Change Facilitator, at firstname.lastname@example.org