Union Bank of the Philippines

The current world is at the cusp of Industry 4.0, embracing digital twins and revolutionizing businesses across the globe by bringing in unlimited possibilities. With every industrial product granted a dynamic digital representation, a convergence of the physical and virtual world, organizations can accurately predict the future of the product by assessing their digital counterparts. Industries such as manufacturing, healthcare, automobile, and banking have implemented this revolutionary technology to unlock potential for development.

In an exclusive interview with CIO Outlook, Adrienne Heinrich, VP, AI CoE Head, Union Bank of the Philippines, has shared her valuable insights on digital twins and the possibilities of their implementation in the banking sector.

How are digital twins helping industry leaders in revamping manufacturing and industrial processes?

Industry 4.0 has embraced digital twins, thriving in innovating industrial and manufacturing applications. Good quality data from sensors installed in equipment are crucial to building a digital twin. And with the prevalence of IoT devices, a large amount of data is available to design a dynamic digital representation of the product. It helps us improve product or service performance as we can select the intervention that gives the best result. Digital twins facilitate innovation as insights from simulations inspire new ideas. This way, we reduce the time needed to release a new product or service.

 Digital twins facilitate innovation as insights from simulations inspire new ideas. This way, we reduce the time needed to release a new product or service

Combining AI with the digital twin concept is another trend. Using AI, we can build digital twins with existing data for a physical model. An alternative method is building a digital model using simulated data, gaining a richer training set. This powerful combination results in improved and efficient decision-making and enhances machine learning models. Therefore, industry leaders save a lot in terms of time and cost upon using digital twins.

Can you summarize the applicability of digital twins? What are some of the problems faced by your clients in seamlessly adopting it in the AI space or industrial manufacturing space?

One challenge people face while applying digital twins is the imbalance of data. If there is an insufficient sample size, we cannot build a good representation of the physical world, resulting in an ineffective model.

 It can sometimes be difficult for digital twins to provide a comprehensive perspective of the situation, as in the case of our business unit, a power plant of the parent corporation under which Union Bank of the Philippines operates. When an accident occurs, it affects numerous pieces of machinery rather than just one. It is not an isolated event but rather a correlated one. That is why it is necessary to develop a holistic digital twin covering the entire plant.

Applying digital twins to model people’s behavior can cause discrepancies in its results as human behavior is dynamic. In the banking sector, this problem arises when dealing with fraud. When a fraudster cheats, the digital twin can track the activity without delay. But once spotted, they can immediately retreat, adapt to the situation, and adopt another channel to commit fraud again. Therefore, the digital twin models need to adapt continuously.

Can you share an instance about the usefulness the applicability of digital twins?

Since I am heading the AI CoE of Union Bank of the Philippines, fraud can be a good example. Let me elaborate on the problem and the way we combat it.

During the pandemic lockdown, the Philippines experienced a 200% rise in phishing scams. Filipinos have a very active social media lifestyle, making them easy prey for scammers. For example, a phisher caught a feel of our communication processes, imitating it. The victims believe they are dealing with a legitimate party. Through SMSs and emails inducing excitement or panic, victims click on the malicious links and enter their credentials which are later misused.

We use digital twins in two ways to combat this fraud - phishing prevention and phishing detection. Phishing prevention uses AI algorithms and the digital twin to identify the customer demographics most vulnerable to phishing. With this knowledge, we can personalize communication and be proactive with our antiphishing campaign. Phishing detection uses digital twins and AI to design and train a model that detects the legitimacy of ongoing transactions.

One benefit of using digital twins is that we can simulate different scenarios and trade-offs since we can choose accurate outcomes before deploying them. Different strategies will have a positive and negative impact on it. For instance, a digital twin with 80% accuracy will help us reach more customers who became victims of phishing attacks. On the other hand, 20% inaccuracy will lead us to call customers whose transaction was legit. Risking their irritation is the trade-off for deploying the phishing prevention campaign.

What advice would you give to the people venturing into AI and digital twins?

I would advise them to grasp the problem they want to solve using the digital twins. It is also crucial for them to understand the space they operate in and the kind of representation they want to provide for equipment. For a holistic presentation, one should have good quality data that includes all the real-world situations they want to solve. If not, they will have to return to the drawing board to discuss where to get more data to get a more accurate description to mirror the real-world scenario.