Hello there,

We’re back in the world of deep tech today, with a startup that wants to carve out a space for itself right in the centre of the brain-controlled tech space.

Scroll down to read all about Cogitat. As usual, PreSeed Now members get the full story.

But first! A couple of things for your radar…

• Ready to found a new startup? The final deadline to apply for Zinc’s sixth venture builder, focused on financial resilience, is just over a week away: [More info]

• In Manchester? Friend of PreSeed Now, David Levine is hosting a ‘Coming to America’ event on Monday morning, about planning for US expansion and fundraising. He’ll be in conversation with Daniel Glazer of Silicon Valley focused law firm Wilson Sonsini: [Sign up]

– Martin

Can Cogitat power the future of brain-controlled tech?

No matter the attention companies like Neuralink, Meta, and Snap-owned NextMind have shone on the field in recent years, brain-computer interfaces still feel like something of a sci-fi dream.

But having a direct connection that allows you to, for example, move an on-screen object by just willing it to move with your brain would be mightily useful in all sorts of contexts, and it feels like an obvious step we should be able to make soon given the amount of publicity various efforts in that direction have received in recent years.

With this in mind, London startup Cogitat–an Imperial College spinout–is working on making it easier for developers to add brain control to their apps and hardware.

The idea is that by providing a high-quality interpreter, the startup will lower the barrier to entry for brain-controlled applications without the need for its customers to have their own proprietary tech for the task.

“We're really trying to connect the mind and machine together,” says co-founder and CEO Allan Ponniah. “Essentially, our aims within the company are to build a core brainwave decoding technology; to allow anybody to convert their thoughts into actions.”

How it works

Rather than get involved with brain implants or other specific ways of discerning intent from your thoughts, Cogitat is working on delivering what it aims to be the best possible decoder for translating brain activity into digital actions.

“Essentially we're building technology to work with anybody, anytime, any device, and any place,” says Ponniah.

Co-founder and CTO, Dimitrios Adamos explains further: “Our decoders get an EEG signal in the input and give out predictions about what happened to the output, and that can translate into a control signal. 

“But because we are device agnostic, and we just need some sensors over the central part of the skull where the motor cortex is, we can work with different vendors even with devices that aren’t in the market yet.”

Cogitat has produced a demo video to show the tech in action:

While VR gaming, ‘metaverse’, and stroke rehabilitation applications are perhaps obvious potential targets for the startup, Ponniah sees potential in simple applications like helping people with physical disabilities navigate app interfaces. And that’s just the start: 

“We think it can be used to control prosthetic limbs, to summon your car, drive drones, drive robots, and various other aspects that it could be used for.”

And Ponniah hopes that by making it easier to work with brain control, Cogitat will help to create new markets and applications that are yet to exist.

Progress so far

Unusually for the CEO of a neurotech startup, Ponniah is actually a plastic surgeon by profession. 

“Originally when I went into medicine, I wanted to be a neurologist,” he says. “But at that time, the technology wasn't really advanced, or the treatments really advanced. At that time, you were able to give somebody a diagnosis with a very long name, and that was about it. They were stuck with it and there were very limited ways you could improve their care.”

In his work in the field of plastic surgery, Ponniah says a lot of his work has focused on craniofacial surgery around the face and brain, touching on similar ground to neurology.

“We have done surgery in the past that improves people's brain function from various facial deformities by improving their skull shape,” he says.

But his professional background and interest in tech led him into the startup world in partnership with Stefanos Zafeiriou from Imperial College.

Then Zafeiriou met Adamos to discuss the latter’s research into using brainwaves to understand an individual’s music taste. This led to a partnership with the Science Museum in London that Adamos says collected almost 800 EEG recordings from museum visitors in a month.

Eventually this project evolved into the beginnings of Cogitat. The startup was formed as a company at the end of 2020, and formally spun out of Imperial College in the summer of 2021. Since then, they’ve submitted patent applications, and they beat the US Army to win a competition at the NeurIPS conference in 2021. 

Cogitat collects EEG data as an ongoing process at their lab at Imperial College. Ponniah claims the startup now has “the largest EEG data set for the motor cortex in the world.” That data is key to the machine learning Cogitat uses to make accurate interpretations of an individual’s brain activity.

But isn’t brain-reading tech… scary?

As impressive as brain-based control of software and hardware can be, some people might have a certain level of teeth-on-edge nervousness about tech that can ‘read your thoughts’.

Given many users see a convincing ‘personality’ in the current generation of conversational A.I. like ChatGPT and interpret that as sentience (when really there is just the illusion of a personality), it’s understandable some might over-egg what Cogitat is doing here. 

“I'm fully aware that there are some negative connotations of looking into people's minds and people are very frightened of that kind of possibility,” says Ponniah. “Our focus is only to convert people's intention into action, rather than looking into the inner workings of their mind.”

Investment, vision and challenges

Dive deeper with essential startup information for PreSeed Now members:

Investment: