DSEI UK 2025: Metis launches THEIA, AI-enabled drone detection

Metis, a leading provider of radio frequency (RF) and drone detection technology, launched a new operating system and user interface, named THEIA, during DSEI UK 2025. The system is designed to enhance the capabilities of the company’s drone detection products by leveraging machine learning (ML) and artificial intelligence (AI).

The new software will be installed on the company’s full product line, including the mobile drone detection system Skyperion, which Metis also show-cased at the event. THEIA, named after the Greek goddess of insight and vision, has been developed to consolidate complex RF data into a single, user-friendly graphical interface. The aim is to provide operators with rapid situational awareness, reduce their cognitive burden, and streamline workflows for faster and more accurate decision-making.

“For many, deciphering the mass of data traditional RF detect systems produce must seem like trying to read the matrix,” Tony Burnell, CEO of Metis, said in a September press release. Traditional approaches to signals intelligence and electronic warfare tend to visualise the raw signals that are detected by the antenna and receiver. In a busy urban environment, this can create an intense and complicated picture that is difficult to understand. 

Electronic warfare operators tend to be highly specialised personnel as a result, with years of training that is difficult to replicate or replace without a similar process. Machine learning is particularly applicable to RF signals because a set of algorithms can be trained on synthetic data that is representative of the electro-magnetic spectrum. This enables them to rapidly make sense of operational data, and through a relatively low number of training cycles, provide a meaningful increase in capability. What this means is that a system like THEIA can simplify the analysis stage, helping to highlight to an operator what is relevant and what is not, for example. 

It will also complement Skyperion, which has been designed to be used with little to no operator training, according to Metis. To some extent, this reflects the fact that drones are a threat and problem to all elements of a force, and they cannot be left to a select few EW specialists. This is reflected in Securis, another project that Metis is contributing to in partnership with Allen Vanguard. 

According to Metis, THEIA is developed to be future-proof, with an open architecture that allows for easy integration with existing command and control systems and future upgrades. The interface can also be customised for operators, from basic to expert levels, depending on their training and the threat profile.

Calibre comment

There is a growing trend to streamline signals processing and intelligence with AI. It can be observed both in sonar operators and in EW, where the skillset to analyse and understand received signals is extremely specialised and in short supply. As mentioned above, this is in theory an area where AI can really help defence. Sonar and RF data can be represented in formats that lend themselves to training an algorithm. And there are often sets of data that are commercially available for the initial training. Some companies from Ukraine are also offering their signal libraries for this very purpose. 

By Sam Cranny-Evans, published on September 17, 2025. Credit for the lead image is Metis.  

If you like what Calibre Defence is doing and would like to support us, please do me a favour and sign up for our newsletter, which you can find on the homepage.