Traditional biopharmaceutical process development relies on experimentation. To understand how various process parameters impact the quality and consistency of the finished drug, scientists conduct hundreds of separate studies and assays.

A more modern approach is to try and automate parts of process development according to Russell Green, director of product application at research automation specialist, Automata.

“Automation is already heavily used in lead discovery and screening but less so elsewhere. Process development has been growing in the automation space, but it is only moving to a high throughput era now and starting to reap the benefits of large dataset collection,” he says.

Biopharma’s growing willingness to embrace automation outside the discovery lab reflects a desire for consistency, both in terms of the product and the data.

“Generally speaking, automation brings reproducibility and throughput. It comes back to the main reasons to automate in life science research—to generate more data of higher quality, faster,” continues Green. “In particular for biopharma, highly contextualized data is important. For example, tracing a complete dataset including all metadata back to individual clones is required for FDA submission for an investigational new drug.”

Demand is on the rise

And demand for automation is increasing, points out Green, who says while some companies see it as a way of increasing efficiency others view it as a foundation for the adoption of even more advanced digital methodologies.

“There are several areas of innovation being asked for depending on what area you are looking at. However, there are common themes. ‘Do more with less’-ways of utilizing both their equipment and space more efficiently. This is an area we are closely focused on,” explains Green. “Then there is data aggregation, structuring and utilization. If we want to train AI models correctly, we need highly characterized data. There has been a reproducibility crisis in life sciences for a long time, partly due to the lack of automation and digitalization. This is something most companies are trying to solve. It is hard to utilize new AI technologies without these deep datasets.”

Personalized medicines

And interest is not limited to firms that make mass market products. Developers of personalized, patient-specific therapies are starting to automate processes.

“An autologous cell therapy is derived at a patient specific level and so the areas to automate really come into QC of prepared therapies. This becomes a much more regulated type of automation as it is part of a manufacturing process,” according to Green. “For the actual culturing of cells for the therapy manufacture, this is much more of a closed system, looking at companies like Ori and Cellares that are innovating in this space for fully closed, personalized medicine manufacture.

“Depending on the modality, there are varying levels of automation in the R&D, process development, and manufacturing spaces.”

Previous articleExplainable AI Directs Small Molecule Structural Class Drug Discovery
Next articleBenchtop Gene-Therapy Analytics Tool Adds New Capabilities