DIY brain-computer interfaces have arrived — why that’s cool (and why it isn’t)
Can a DIY BCI powered by a Raspberry Pi actually be worth a try?
A US-based science group intends to use a “crowd sale” to distribute brain-computer interfaces to a community of hackers. Dun-dun-dunnnnn!
Another way of putting it: A PhD student and an artificial intelligence expert are planning to sell some nifty brain-computer interfaces they cobbled together. And, for some weird reason, they’ve called their company Hacker BCI and given it the tagline: “from hackers for hackers.”
Per the duo’s website:
In our definition “Hackers” are creative problem-solvers, who will find something that others won’t be able to find, and think the way others won’t think.
The NSA might disagree.
Up front: Brain-computer interfaces are huge right now. Or, at least, they could be — if they were accessible, affordable, and functional.
There’s enough hype out there to make “BCI” a household acronym. And we can thank Elon Musk’s Neuralink for that.
Neuralink had raised over $200 million in funding as of about eight months ago. With that money, Musk’s genius, and a revolving door of top-notch talent, they’ve managed to show a video of a monkey playing Pong with its brain and another where a computer makes a beeping sound as a pig snuffles.
While those are very impressive use-cases for a device that’s surgically implanted in your head by robots, the price is a bit steep for most of us.
Background: Brain-computer interfaces have been around for decades. But, thanks to recent advances in machine learning, they’ve become a hugely popular field of research-for-future-profit.
There’s a huge difference between the wearable and intracranial interfaces used in the medical community as diagnostic or assistance devices and those being developed for everyday use by the general public.
What Hacker BCI is attempting to accomplish is quite different than what the medical community or Elon Musk is.
Per a recent article on Hackster.IO:
Maker Ildar Rakhmatulin is working to make entry-level brain-computer interfaces affordable and easy to build, designing an open source add-on for the Raspberry Pi family of single-board computers to interface with electroencephalograph sensors: the PIEEG.
For more, you can check out this weird YouTube video:
Our take: This is a mixed bag. On the one hand, it’s super cool. The company’s paper says it costs about “$350 for 24 electrodes.” We’re assuming that’s indicative of their cost of production for the devices.
This makes it seem like you could build your own DIY BCI for around $500 or $600. Nice.
But, on the other hand, this isn’t the first BCI kit for hobbyists and students. It might be the first one with a bespoke Raspberry Pi shield and a research paper. But a quick Google search can point would-be BCI builders to myriad similar solutions with options galore.
The bottom line is that these devices can only do so much. Sensing EEG, EMG, and EKG through a human skull and the tissue and hair covering it doesn’t provide high resolution data.
At their absolute best, these sorts of BCIs allow you to control computers with movement. You put on a cap, and a computer can tell when you intentionally blink, move your eyes in a specific direction, and clench your jaw.
Clever algorithms can translate these signals into direct inputs, such as executing a function whenever you blink.
Is it worth hundreds of dollars? In the practical sense: no. Of course it isn’t. You’re essentially paying money so you can sit in a chair with a bunch of wires hanging off your head that make it more difficult to perform simple computer functions than using a keyboard, mouse, or touch control.
There’s no functional, consumer-luring upside here. The reason you purchase a low-end BCI project kit is for the joy of tinkering or as a teaching tool.