Brain-computer interfaces have been a staple of hacker projects for years — reading raw EEG signals, blinking LEDs with your brain, maybe moving a servo if you're feeling ambitious. What's been harder is making something useful with those signals in a real application. The team at PiEEG is taking a run at that gap with PiEEG XR, an open-source neural face interface designed to snap onto VR headsets and feed brainwave data into spatial computing environments.

The hardware is built around their IronBCI board: 24-bit ADC resolution, 250 samples per second, BLE5 for wireless data transfer, and a low-noise analog front end. If you've worked with cheaper EEG setups before, you'll appreciate why those specs matter — most affordable BCI hardware drowns signal in noise before it gets anywhere near your software stack.

The Architecture

The system splits into two repos, both open on GitHub:

• PiEEG-server — the backend server that ingests raw EEG streams, handles signal processing, and exposes a clean API downstream

The server handles the heavy lifting: artifact rejection, feature extraction, and classification. Downstream of that you get a Focus-to-Action API — essentially a real-time stream of mental state data (concentration level, emotional valence) that game engines and XR environments can subscribe to.

The physical interface snaps onto a VR headset and positions the electrodes against the face, picking up frontal lobe activity. No gel, no cap, no 20-minute setup ritual.

What You Can Build With It

The two flagship demo use cases are:

Affective avatars — the avatar mirrors the user's emotional state in real time. Joy, stress, focus — the system reads neural signals and maps them to avatar expressions and behaviours. Useful if you're building social VR or telepresence applications and want something more expressive than tracked face muscles alone.

Focus-to-Action — sustained concentration becomes a control input. In the existing demos, a user can stare at a virtual object, concentrate, and cause it to react (melt, explode, transform) based on brainwave intensity. Think of it as a hands-free interaction channel — potentially interesting for accessibility applications, or just for building the telekinesis mechanic you always wanted.

Both are exposed through the API, so you're not locked into their demo environments. Roll your own Unity scene, hook it up to Godot, pipe it into whatever you're building.

Demo 

Blinking example 

Follow us on Kickstarter

or just drop your email here https://xr.pieeg.com/, we will notify you when PiEEG XR will be live