All in the eyes: How a Dubai-based deep tech company is empowering its scientists, engineers, and researchers with AI to develop next-generation smart contact lenses
In a brightly lit lab in Dubai, filled with the gentle hum of countless PCs and printing stations, Valentyn Volkov carefully lifts a stainless-steel stand to peer through the single contact lens held upright at its end. At first glance, the almost invisible lens – barely 15 millimeters in diameter – looks no different from the clear contact lenses found in optical outlets across the city. It is only when you look through it, that you see the tiny spaceship hologram that pops up in your line of sight.
“Space Invaders,” Volkov confirms with a small smile – most people recognize it almost instantly. Gesturing to the prototype, the co-founder and chief technical officer of XPANCEO, explains that this is just one of the many applications of the next-generation smart contact lenses that the deep tech company is developing. Imagine watching a live football match on the public transport, casually gaming with friends during your lunch break, making instant payments while shopping online, having real-time text translations, health alerts and medical reminders when it is time to refill your prescription – all this via an invisible and weightless smart contact lens that is as natural to the wearer as their own vision.
XPANCEO’s vision is for these next generation smart contact lenses to one day replace all the smart gadgets that we are using in our daily lives like our smartphones, smart watches, and fitness trackers – to have an “infinite screen” that is the perfect interface between the real and digital world, and which gives users an ultimate experience.
But why contact lenses? “A contact lens is a unique device because they are extremely slim, extremely compact and almost invisible; once we put them on, they are a part of our body and we forget about them,” Volkov explains.
A challenge on the nano scale
Working with contact lenses, though, comes with its own set of challenges. “Trying to integrate all the optoelectronic components inside such a delicate and compact device that is less than one millimeter in thickness is extremely difficult,” says Volkov. “Also, you can’t use traditional materials as they simply don’t work at this scale; so you have to use new cutting-edge nanophotonic materials.”
The issue with these nanophotonic materials is their novel nature; many of those compounds are not in the market at all, rather, they were discovered only a couple of years ago, not enough time, Volkov says, for scientists and researchers to fully understand their unique properties. At the same time, even materials that have been well known in their bulk form for decades remain largely unexplored in terms of their optoelectronic properties when reduced to low-dimensional structures. Yet, it is precisely these low-dimensional materials that hold the greatest promise for building the optoelectronic components required for XPANCEO’s smart contact lenses.
“It’s like working with a completely new periodic table of chemical elements, many of them absolutely unknown to us because they only appeared on the scene a few days or weeks ago,” Volkov notes. “How can we know which of them are good for and which are not suitable for our contact lenses? The only way to find out is to carefully study each and every one. By running material science and optic experiments, accumulating the results, and then comparing them to previous studies to determine which materials and their specific combinations would be ideal for our product.”
This painstaking process often ends up taking several years, a time frame that is inconceivable for a company that hopes to deliver its flagship product as soon as possible. With traditional methods out the window, Volkov says that the company turned to Artificial Intelligence to help accelerate its research and development cycle.
AI as a research partner
With AI, XPANCEO’s teams were able to not only shrink all their R&D procedures from years into just months or even days but also anticipate material properties and identify optimal candidates for integration into their products. “AI gave us a unique opportunity to create new non-existing materials with electronic and optical properties on demand and that has been absolutely invaluable, giving us clear hope that we can do our project within the time frame of our roadmap,” Volkov says.
Like many in the scientific community today, Volkov says that the time for AI experimentation is over and that organizations in highly competitive fields have moved on to active implementation of the technology. XPANCEO’s AI strategy was to develop its own AI platforms: The first was an Artificial Intelligence Materials Hub, which brings together digital tools as well as information about material critical structure predictions. The second is a system called PRISMA – Patent Retrieval and Intelligent Scientific Multidimensional Analytics – designed for rapid analysis of large volumes of patents and automated generation of structured reports in just a few clicks. The results of using both platforms have been stellar.
“Previously, our patent check used to take us days but now we can do it in a few hours, and we went from analyzing hundreds of documents to analyzing thousands within the same time frame, so we consider this to be amazing progress. It is almost like a superpower, once you start working with AI, there really is no going back.”
A learning curve that requires regular AI skilling
Looking back at XPANCEO’s AI adoption journey, Volkov admits that the process was not without hurdles. Early on, researchers at the company realized that there were issues with the reliability of the data obtained. Having chosen five of the most promising AI tools that were available in the market at that point in time, the researchers gave them each a task to find specific data from various scientific papers. “We found that some of the AI solutions very convincingly gave us information which was not present in the scientific papers at all, so that was one of the drawbacks of the system. We also checked to see how efficiently they could analyze scientific materials. Once we started to play around with them, we realized that sometimes they work well. For example, we had a chatbot find a patent which was missed by our experts, so that was great,” Volkov recalls.
Another issue that the company quickly identified was that early enthusiasm for AI had resulted in employees adopting several different AI tools and solutions to help with their daily workloads – a scenario that could lead to serious consequences, Volkov explains. “The danger of not using the right AI tools is that you will lose the direction of your research, you will lose time, you will lose money. And, in the end, you will lose advantage in this highly competitive landscape where you and other companies are trying to deliver their product to the market as soon as possible.
“The danger of not using the right AI tools is that you will lose the direction of your research, you will lose time, you will lose money. And, in the end, you will lose advantage in this highly competitive landscape where you and other companies are trying to deliver their product to the market as soon as possible.”
XPANCEO’s solution was to invite a team of dedicated AI experts to act as an advisory board. These experts were tasked with cutting through the chaos of early adoption and helping skill the company’s teams on how to effectively and efficiently use the right AI tools. “Our team is made up of people from different fields like research, engineering, and health, all working together, so we needed to ensure that everyone was on the same page when it comes to AI,” says Volkov.
The team of AI experts quickly launched a series of online seminars with live Q&A sessions for all employees, in addition to regular newsletters which talk about new AI tools and features, how to use them, and even a guide with best practices so that everyone in the company can learn and keep up with the latest developments in AI technology. Their diligent efforts paid off, and XPANCEO’s R&D efficiency doubled over the last six months, Volkov revealed. “Our scientists are now able to focus more on their actual work, knowing that there is someone supporting them with the right AI tools. Collaboration has also improved, and more and more team members are becoming what we call ‘AI champions’ – they’re helping others to use these tools more efficiently – and we are very proud of this.”
Picking the right partner to tackle bottlenecks and ensuring security
Beyond investing in a dedicated team to lead AI skilling efforts across the company, Volkov also stressed the importance of having the right partner and tools at your side to scale operations and ensure security. “Due to the specific nature of our work, where a lot of IP and private information is involved, it is very crucial for us to work with a partner like Microsoft which can provide us with scalability, security and efficiency.”
Volkov noted that Microsoft’s technologies are deeply embedded in XPANCEO’s AI workflows. “We use a range of Microsoft tools, including Azure OpenAI API for analyzing research papers and patents. All our computational power is hosted on Azure, which is essential given the level of security our AI tools require to examine the properties of advanced materials.”
He added that Microsoft’s AI tools have helped the team overcome major bottlenecks in their processes. “Previously, we spent days manually reviewing and analyzing new patents and research papers. Now, with Microsoft’s solutions, we can complete these tasks in just a few hours. That’s a game-changer – it significantly boosts our efficiency and brings us closer to our ultimate goals.”
In addition to technology integration, Microsoft is also playing a role in upskilling XPANCEO’s workforce. One of the company’s research scientists recently completed the Microsoft Azure Fundamentals course, successfully passing the certification exam after rigorous study. The experience not only deepened his understanding of cloud technology but also proved invaluable for both work-related projects and personal development. He particularly appreciated the strict proctoring measures, which ensured a high standard in the certification process.
“Due to the specific nature of our work, where a lot of IP and private information is involved, it is very crucial for us to work with a partner like Microsoft which can provide us with scalability, security and efficiency.”
A smarter, faster, AI-powered future for science
Looking ahead, Volkov is quick to point out that AI should not be seen as a magic solution, but rather as an important tool – powerful, precise, and always handled with care. “Our approach is to treat AI like any other lab equipment. We’re scientists first and foremost, and while AI can predict material properties, identify patterns, or analyze research papers, every output is still reviewed and validated by a human expert. This double-checking is a fundamental part of our process. Also, AI will never replace scientists; it will work alongside them, enhancing their capabilities and accelerating the pace of discovery.
In many ways, the whole system of AI in materials science is still in its early stages, Volkov notes. “But that’s exactly what makes it so exciting. There’s a real opportunity now for AI and modern R&D professionals to evolve together.”
For Volkov and his team, success in this new era of scientific exploration will depend on more than just data or AI tools. “The main requirements for any scientist working with AI today are curiosity, patience, and bravery,” he says. “With those, we believe research teams can take scientific discovery to a whole new level.”