Doubling Down on Innovation With a Digital Twin

Approximately half of new technologies fail to reach the market, and the actual number is probably higher, according to Leyla Mirmomen, founder and CEO of Optimuos, a client-based engineering company headquartered in Los Angeles.

From insufficient integration of customer needs into a specific technology and regulatory complications to prolonged and intricate development processes, most market failures stem from roadblocks that are inherent to conventional approaches. Traditional development methods tend to worsen the odds of launch because they delay communication of customer feedback and market insights. An alternative method exists, says Mirmomen, that can help developers to improve their chances of accelerating time to market and achieving a timely product launch.

The trick is "commencing an exhaustive study of customer requirements in tandem with the initiation of the development and primary design phases," said Mirmomen. "This early-stage exploration facilitates a comprehensive understanding of engineering challenges, cost implications, and regulatory constraints. Armed with this knowledge, early design iterations can be implemented, reducing the need for substantial investments. This strategic approach significantly improves the likelihood of formulating a technology solution that aligns with market demands and identifies potential failure points from the outset, enabling a shift toward more viable alternatives."

Key to achieving development goals for new technologies is the utilization of what’s known as a "digital twin" — a virtual version of a real device that’s generated through a blend of data and physics. Falling within the realm of artificial intelligence (AI), digital twins can reportedly serve as a valuable component to enhance decision-making by providing a system’s virtual representation. Mirmomen and her colleague Pirouz Kavehpour, a professor of engineering at UCLA, presented an educational session on digital twins at the Medical Design & Manufacturing (MD&M) West event in Anaheim, CA, running from Feb. 6 to 8. During the session, Mirmomen and Kavehpour discussed the sequential steps required for constructing a digital twin while exploring various applications and its potential as an AI-driven predictive tool.

Driven by data

Validating the potential of new technologies using a digital twin melds data-driven insights and core physical principles to craft predictive models that serve as agile virtual systems, says Mirmomen. "In the current landscape, data play a pivotal role across diverse applications, encompassing a broad spectrum of processes under the umbrella term 'data driven,'" she said. "Our organization strategically employs data-driven insights in two primary capacities: We leverage data to develop highly accurate and cost-effective simulations of systems, and we utilize data to construct various scenarios, subjecting the digitally engineered system to these scenarios to iteratively enhance system behavior and functionality." These methodologies produce multi-layered benefits, including alignment of technologies with targeted objectives and precise predictions of system behavior, Mirmomen said. "From the initial stages of concept development, it's imperative to be conscious of the standards and regulations with which the device must be in compliance," she continued. "Having this data at the outset allows us to identify potential technical challenges and streamline the regulatory path. For instance, designing a device using off-the-shelf parts that already comply with standards expedites the regulatory process."

Contrary to conventional methods, where product design is the first issue of focus while regulatory considerations follow post-prototype testing, digital twinning advocates for integrating regulatory steps earlier in the concept-development phase. "In today's dynamic market, while meeting all the necessary standards is crucial, the most significant element, alongside compliance, is the ability to swiftly enter the market with a reasonable investment," Mirmomen said. "Leveraging technology is key to designing a device that not only aligns with required standards and regulations, but also anticipates future market needs. It's essential to capture and implement customer feedback rapidly, ensuring a competitive edge and responsiveness to evolving market demands."



Going beyond AI


While AI tools require algorithms to make decisions based on the events they are trained for, a digital twin provides a virtual representation of the system it is trained for and can serve as a more valuable component to enhancing decision-making, says Mirmomen. 

"The integration of a digital twin into AI decision-making processes can significantly decrease the initial dependency on a massive amount of data," she said. "The virtual representation offered by the digital twin enables more accurate decision-making without relying solely on extensive datasets. Moreover, the digital twin exhibits a unique capability to interface seamlessly with the physical device. Through constant learning and iterative improvements, it attains the proficiency to mimic the exact behavior of the system in real time. This transformative capability positions the digital twin as an invaluable asset, not only in predicting system behavior, but also in actively adapting and evolving alongside the real-world counterpart." At the same time, the use of digital twins aligns seamlessly with AI to ensure smarter, precise solutions and bridges the gap between ideas and reality for agile development, meticulous validation, and streamlined progress, according to Mirmomen.


Comprehensive collaboration

As technology continues to advance, there’s evidence of a "fifth industrial revolution." With this as a backdrop, there is a discernible trend across diverse industries, including medtech companies, toward the integration of digitization and related technologies within their research, development, and engineering frameworks, says Mirmomen.

"In the swiftly progressing era of digitization, where technological advancements are pervasive, product development emerges as a sector ripe for substantial digital transformation. I firmly believe that understanding market needs and knowing your customers, along with the regulatory and standards landscape, should shape the primary design of the device. This awareness should extend through development and prototype testing until the device is ready to launch. This approach ensures compliance, user-friendliness, and cost-effectiveness."

At Optimuos, companies are engaged across various project stages spanning from initial concept development to the optimization of fully functional devices already in the market. Serving as an extended research and development arm for clients, Mirmomen and her team adopt a collaborative approach throughout the development journey to concurrently constructing the virtual system while actively contributing to the product’s design, testing, and prototyping processes.

"We optimize the utilization of available resources, ensuring a synchronized and efficient development process," said Mirmomen. "This collaborative method not only facilitates the timely achievement of goals, but also promotes cost-effectiveness by leveraging existing resources effectively."

De-risking technology through simulation

Another crucial element to leveraging the benefits of digital twins is the "in silico,” or virtual, trial that aligns with the FDA-recognized use of simulation in the regulatory process. This approach de-risks technologies through simulation — by mimicking the physical system virtually and significantly reducing the size of clinical trials, thereby impacting both the cost and time associated with the regulatory process. "The digital twin plays a pivotal role in this concept by providing a virtual representation that behaves like the physical counterpart," said Mirmomen. "This enables our team to run diverse scenarios, integrate patient data, and record outputs, facilitating continuous improvement of the device while capturing valuable insights."

Mirmomen believes the use of a digital twin can become the core to any decision-making for future AI tools, simplifying and enhancing their accuracy. "The paramount aspect of digital twinning lies in its capability to ensure simulations accurately capture the fundamental physics of the device," she said. "Leveraging advanced methods for refining and implementing data, the digital representation of the system closely mirrors the authentic functionality of the device. This integration not only elevates the efficiency of product preparation for production, but also plays a pivotal role in pushing the frontiers of innovation within the medtech industry."