During a discussion on Industry 4.0, Andy Topping, CSO at Fujifilm Diosynth Biotechnologies, outlined that the industry needs to undergo a cultural shift to better generate and utilize data gathered from equipment, during UK BioIndustry Association’s annual conference at the end of last year.
Topping stated that the general aim for Industry 4.0 is to take ‘individual operations’ and then ‘connect them all together’.
“So, instead of having a process which involves 10 separate, computer-controlled things, it becomes one entity in its own right. It’s basically Skynet, without the destruction of the human race,” he added.
Comparing the state of the bioindustry to other industries, Topping outlined that a modern jet engine contains around 5,000 sensors providing constant data on performance.
For his part, Topping described how the automotive industry produces vehicles that are “much more sophisticated and generate much more data” when compared with bioprocessing equipment.
Fujifilm is not the only company looking at other industries to determine the efficiencies and advantages of Industry 4.0 technologies. GSK has previously described how it invited other companies, including those from the gaming industry, to explain how it had implemented such technologies.
When analyzing the capability of bioprocessing equipment in comparison to other industries, Topping concluded that it was “pretty dumb.”
“In a typical chromatography skid, it basically has two pressure sensors on it – and they’re only there as safety controls to stop us blowing it up, not to control the process. We have a flow meter, but flow is one of our inputs. We have two conductivity meters and, typically, a UV detector. That’s it,” he described.
Topping continued to say that any data generated has to pass through proprietary software, which often just generates PDF documents, with a limited ability to work with the real data.
Working without data
As a result of the data being held within the equipment, Topping suggested that the industry is suffering from the lack of data generated.
One of the problems the lack of accessible data creates is that other key elements of Industry 4.0, such as machine learning and AI, cannot be implemented to the same scale as other industries.
“As much as I love to hear about AI and machine learning, as well as the work being done to drive down the amount data points needed to put in, I think you need to give it something because we’re really giving it next to nothing.
“Even the best AI is going to struggle with next to no data. There’s a big deficit, in terms of what the manufacturing equipment is actually telling us,” stressed Topping.
Building systems from the ground up
With equipment not providing the data required to fully implement Industry 4.0 technologies, Topping described how Fujifilm Diosynth Biotechnologies had reacted by creating its own pilot continuous process at production scale, with a 500L perfusion bioreactor.
The company then worked with vendors and partners to create a completely new, fully connected system, as a necessary step in creating a fully continuous process.
“It’s too hard to build this [system] from the bits and pieces [of equipment] available at the moment – it’s easier to just build a new one. The system has been custom built, which is able to run all the pieces of equipment and because it’s the same skid that runs all of the operations, we can actually link them together and allow them to talk to each other,” Topping described.