Machine learning and AI are widely applied to a variety of different pattern recognition problems on a daily basis – from facial recognition to fraud detection. But did you know it can help decode brain waves?
Cogitat, part of NATO DIANA’s 2025 cohort of companies, is a spin out from Imperial College London’s Department of Computing, based in the UK. Their technology uses advanced machine learning and wearable sensors to decode brain signals – which can be of use in many areas, from supporting rehabilitation by retraining brain pathways, to monitoring cognitive load and fatigue.
While to the untrained eye the electrical signals produced by our brains may seem hard to unravel, there are patterns that correspond to different intentions and cognitive states, which machine learning can learn to recognise. This is what the team at Cogitat are beginning to translate:
“We started by taking a technology known as EEG, which is present in hospitals and labs. This consists of electrodes that you can connect to someone's head to measure their brain waves,” explains Allan Ponniah, Cofounder and CEO at Cogitat.
“However, this technology does not know what those brain waves mean. So we decided to use machine learning to identify the common signals and patterns amongst people, so that we could then put a headset onto anybody and it would be able to read their specific brain waves on a specific task,” he added.
One of the first applications of Cogitat’s technology was stroke rehabilitation. As part of this effort, the team developed VR games that respond to patients’ thoughts as they imagine moving their limbs – even if the muscles don’t respond – helping to retrain brain pathways.
Allan Ponniah and Dimitrios Adamos, Cofounders of Cogitat.
Real-time cognitive monitoring to track focus
Within the DIANA programme, the team at Cogitat are using their tech for cognitive monitoring in the military. Because their solution can track whether someone is focused or distracted, it has potential in military training – by being able to improve the selection of the best individuals for specific critical roles where sharper focus is needed.
“What we did through DIANA was taking our technology and adapting it so it could be used for training pilots. We even went to the Estonian Aviation Academy - one of DIANA’s test centres – and collaborated with them on testing this technology with educators there,” said Dimitrios Adamos, Cofounder and CTO at Cogitat.
Using their solution, which features a software decoder in a headset – worn like a headband – to read brain signals, the team at Cogitat demonstrated that they could distinguish between trained pilots and novices by the differences in their brain waves.
“In a nutshell, this technology can help us find the best minds for high-stakes situations, and train them better and faster. Our goal is to build a system that allows them to perform at optimal levels by tracking their alertness and focus – and alerting them if their attention starts to slip,” said Adamos.
Cogitat's CogiSense platform demo, featuring DIANA TEVV validation.
The value of end-user feedback
Before entering the DIANA programme, the team had only briefly considered a dual-use strategy for their solution and business:
“Our discussions with the military were short before joining DIANA. The initial guidance we were given was to present our final product to them, and they would then assess whether it was needed or not,” explained Ponniah.
“However, DIANA gave us the opportunity to take a completely different approach, where we were able to directly engage with end-users and actually find out what they needed, and what problems they faced – making the most of end-user feedback early on and allowing us to create this product that they can now benefit from,” he added.
Successes during the DIANA programme
While testing their solution at the Estonian Aviation Academy, Cogitat also developed a connection to a defence prime, which they hope to continue to work with in coming months.
“We got into DIANA with a low TRL of 4, and as we went through a customer discovery process during the programme, we developed a better understanding of what we needed to focus on, which allowed us to reach a TRL of 7 and test that within the test centre in Estonia.
“Before going into the DIANA programme, I personally didn’t think that type of rapid progress, especially in the defence landscape, was possible,” said Adamos.
Technology readiness level, or TRL, indicates how developed a technology is on a scale from 1 to 9. This helps assess whether a solution is at the idea stage, or close to real-world use. This helps investors and partners understand that the technology is not just promising – but reliable and able to make a difference in real-world scenarios.
“We went from a generalised technology that could be applied in many scenarios without calibration and on any device, to a completely different testing environment involving an actual flight simulator with real pilots. During this exercise, we were able to provide real-time analytics, so pilots were able to watch the video of the performance during the simulation and understand where they lost focus.”
What is next for Cogitat?
Since finishing the programme, the team at Cogitat have started testing their technology to improve surgical training as well – highlighting the dual-use focus behind DIANA’s mission:
“Previously, the civilian use for our technology was in stroke rehabilitation. But after working on improving pilot training during the DIANA programme, we're now applying what we’ve learnt to surgical training, which is a new civilian setting for our tech – highlighting how the dual-use aspect seems to work both ways for us,” said Ponniah.
Cogitat’s next steps will also focus on deploying their technology via DIANA’s Rapid Adoption Service – a regulatory framework enabling DIANA to award R&D and prototype contracts on behalf of Allies and NATO communities to support rapid co-development, iteration and integration of innovative tech – and on building new connections as they continue to engage through DIANA’s alumni network.
“The customer discovery process that we have gone through as part of DIANA has also opened up discussions with different defence and military organisations around the world, so we will continue to engage with them in the coming weeks and months,” said Ponniah. “We are excited to scale our solution and see it being deployed in real-world environments where it can make a difference.”