Bringing digital twins to boost pharmaceutical manufacturing

Digital twins — a digital replica of the factory floor — are an important part of the rapid digital transformation of traditional manufacturing known as Industry 4.0 So what about Pharma 4.0?

Pharmaceutical manufacturers are increasingly interested in the tenets of Industry 4.0, including the use of digital twins to simulate, test and optimize manufacturing processes on a computer before using them in production, according to technology advisory firm ABI Research. It projects spending by pharmaceutical manufacturers on data analytics tools—including the digital twin — to grow by 27% over the next seven years, to reach $1.2 billion in 2030.

As with other manufacturers, pharmaceutical makers plan to use the digital tools to boost productivity and to track their operations. 

SaaS AI

Toronto-based Basetwo recently moved into this market with its software-as-a-service (SaaS) artificial intelligence (AI) platform. Today, the year-old company announced an upcoming $3.8 million seed financing round led by Glasswing Ventures and Argon Ventures.

“When you have a digital twin, you can use it to run all sorts of scenarios. Like, if I change this lever or knob, how does that impact yield efficiencies?” said Thouheed Abdul Gaffoor, the company’s cofounder and chief executive officer. “You can uncover the best ways to operate the bioreactors and chromatography columns as opposed to doing it with actual prototypes.”

McKinsey notes that the type of analytics digital twins provide, in conjunction with other Industry 4.0 technologies like robotics and automation, typically boost productivity for pharmaceutical manufacturers by between 50 and 100%.  Average-performing facilities could see improvements of 150 to 200%, according to the management-consulting firm.

The Basetwo no-code software platform uses AI to find and learn correlations between various data points returned by the digitally simulated scenario.

In the pharmaceutical industry, bioreactors are used in the industry to produce compounds and substances with the help of cells or whole organisms. These compounds are then used as finished products or undergo additional processing steps to get an isolated compound, such as vaccines or proteins. Chromatography columns separate chemical compounds. Both are vital—and expensive—pieces of equipment that can afford no downtime and that must operate efficiently for maximum cost savings.

But downtime can happen when, for example, bioreactors can become contaminated through improper maintenance, failing equipment or if feeding ports aren’t sterilized long enough or at high enough temperatures. Sometimes gaskets or O-rings are missing or improperly sealed.

For this reason, digital twins are also used to track maintenance and operation issues and to identify parts that are close to failing, so they can be replaced soon. For that, the digital twin depicts actual operating conditions, thanks to information continuously received from the many sensors on the equipment via the Internet of Things (IoT),

Use cases spanning industries

Gaffoor offered another example – a pharmaceutical company that makes biological therapies for rheumatoid arthritis or cancer treatment. The protein for such a therapy would be produced in a bioreactor.

“An engineer can use our platform to pull data from the bioreactor and build a simulation model of the bioreactor and a digital twin,” he said. “Then, the engineer can do experiments like, ‘what if I change the temperature or cell culture pH?’ How will that increase the yield of that protein?” Gaffoor said.

The platform can also simulate how equipment like a bioreactor will work with downstream tools such as a filtration system, he added.

Though the Basetwo tool was developed for pharmaceutical companies, it can be used in adjacent industries such as chemicals, food and beverage and consumer goods, Gaffoor said.

Larger companies, like Atos-Siemens, make AI-powered digital twin platforms for the pharmaceutical industry. Dassault Systèmes also makes the software. In both cases, the platforms are not concentrated specifically on the pharmaceutical industry, but can be tailored to that application.

One reason for the movement of digital twins into the pharmaceutical’s space is that the number of FDA new drug approvals has steadily increased in the past two decades. Between 2000 and 2008 the agency approved 209 new drugs; that number jumped to 302 between 2009 and 2017. 

“There’s been so much investment in research and development, but investment [in analysis software] hasn’t caught up,” Gaffoor said. 

He hopes to help address this gap with Basetwo’s cost-saving analysis software.

The AI platform can also aid with the labor shortages that have affected the entire manufacturing industry over the past two years. Before AI platforms such as Basetwo were available, highly skilled process engineers and operators would manually track how the equipment was functioning. They would use older software or even paper-based techniques to determine how best to improve efficiency and production.

The no-code Basetwo platform, on the other hand, requires few programming skills to quickly build and deploy digital models. 

Basetwo will use the seed financing to hire data scientists and platform engineers to accelerate platform development, Gaffoor said.