Digital Twins: Reshaping the Pharmaceutical Sector

We often look closely at new technologies that will upgrade how we do things in the pharmaceutical sector—artificial intelligence (AI), machine learning and blockchain technology. By now, you’ve probably read a dozen articles on AI and its potential revolution in the pharmaceutical industry. But, in doing so, we often overlook older technologies that are still relevant and under-utilised. One of the most important is digital twins.

What Are Digital Twins?

A digital twin is a set of adaptive models of a real-world product, system or process. The digital twin emulates the behaviour of this real-world object and is used for simulation, analysis and optimisation. These adaptive models are continuously fed with recent data to make the generated simulation as accurate to the real-life counterpart as possible.

The first digital twin was developed by the National Aeronautics and Space Administration (NASA) in the 1960s, but they didn’t name it as such. This digital twin was used to model Apollo missions and to simulate and learn about the failure of Apollo 13’s oxygen tank. Since then, digital twins have been slowly integrated into several industries, including the pharmaceutical sector.

Digital twin integration hasn’t attracted as much attention as AI. Why? Your guess is as good as ours! However, digital twins have good potential for application in the pharmaceutical industry.

How to use a Digital Twin?

In essence, a digital twin is a simulation of a real-world object or system. Hence, it can be used in almost all pharmaceutical processes. Here are some of the major areas where digital twins can be used:

Pharma manufacturing

Digital twins have the most potential in pharma manufacturing—particularly in helping pharma companies scale up their operations.

Consider the scenario: Your scientists have successfully developed the manufacturing process for a new drug, and you’ve gotten the necessary approvals. Now, you only need to start production. But, to do so, you need accurate diagrams and parameters for a large-scale plant. Obtaining these requires trial and error and time, jeopardising the exclusivity of your product.

Instead, you can create a simulation (or a digital twin!) of the process using real-world equipment parameters and optimise it. Doing so removes all the kinks in the system so that the installation will be simple and swift!

This is the most beneficial application of digital twins. They can also be used for real-time optimisation of existing processes and predictive maintenance, allowing pharma companies to optimise their manufacturing processes beforehand.

Clinical trials

Clinical trials are a necessary, expensive process, and pharma companies have to deal with the uncertainty of volunteers unexpectedly dropping out. Instead, digital twins can be used. Digital twins have two distinct applications in this field:

• As virtual trials: A virtual clinical trial can be performed using digital twins instead of physical clinical trials. While implementable only in some stages of clinical trials, virtual trials can speed up the clinical trial process and reduce human testing.
• As patient twins: Digital twins of patients can also be created based on volunteer profiles. Companies can use these personas for clinical trials instead of performing clinical trials on people.

From a pharma company’s perspective, these applications of digital twins can reduce the dependence and uncertainty of clinical trials, allowing it to utilise its resources better.

Personalised medicine and remote patient monitoring

Like clinical trials, patient twins can also be used to study disease progression in patients and develop personalised medicines. These patient twins can also be used to study the patient’s response to medication and determine possible side effects. For example, patient twins can be used to optimise the dosage for a cancer patient.

In addition, by combining these models with real-life health monitoring systems like smartwatches, the models can predict adverse events. For example, by obtaining real-time data from a volunteer’s smartwatch, the patient twin can track glucose levels and alert responsible authorities before an adverse event occurs.

Such use of digital twins can help pharma companies reduce expenses and better utilise available resources. Many Big Pharma companies like Pfizer and Sanofi have implemented digital twins. However, considering how old the technology is, its application is relatively behind.

What’s Limiting Its Use?

Data limitations

Like AI, digital twins also need high-quality real-time data to simulate an object accurately. However, the data environment in the pharma industry is quite scattered: data is obtained from diverse sets and has different qualities, reliabilities and standards. Hence, integrating all the various data into a standard format that is compatible with the legacy data infrastructure in most pharma companies is challenging.

Financial limitations

To put it simply, developing models that can simulate real-world objects is expensive. In addition, the return on this investment is quite uncertain, which makes many pharma companies hesitant to implement digital twins. Small and mid-sized pharma companies may struggle to justify digital twin adoption because of the high upfront implementation costs.

Regulatory limitations

There are no standardised guidelines for digital twins. Hence, digital twin implementation is quite tricky for pharma companies as they can’t simply define whether a change will result in compliance issues. Many companies are unwilling to adopt digital twins on these grounds. For others willing to embrace, regulators may require rigorous validation of digital twin models before approval of their use, which means higher costs.

Cybersecurity threats

In the pharmaceutical space, digital twins need to use patient data to function appropriately. However, this may bring regulatory scrutiny on the company. Furthermore, it may also make the models a target of cyberattacks and hacking attempts, which are additional challenges that companies must address before implementing digital twins.

While these sound like enormous challenges, historically, the pharmaceutical industry has been able to overcome much larger ones. So, we can expect that despite its slow adoption rate, digital twins will someday be implemented in most pharma companies.

Conclusion

We believe that digital twins are a technology with exceptional potential for transforming the pharmaceutical industry. However, at the moment, it is severely under-implemented and under-utilised. Its implementation can considerably simplify processes for companies; for example, instead of first building a facility and encountering problems, companies can create a perfectly running simulation and then make it. However, to do so, companies will have to replace legacy systems and obtain high-quality, real-world data—two challenges with significant financial implementations.

FAQs

1. Digital twins are commonly used in which industries?

Digital twins are commonly used in the aerospace and automotive industries. However, they have good potential in the manufacturing, energy, construction, and healthcare industries.

2. How long does it take to implement digital twins?

The implementation of digital twins depends on the complexity of the models and objects. Simple models may take up to several weeks. However, complex models may take months or even years.

3. How much does it cost to develop digital twins?

The development and implementation costs vary from one company to another because they include the cost of developing the model + implementing the model + training employees + updating infrastructure + gaining necessary licenses + setting up data security measures.