A Mind-Blowing Levitation Hack

It’s mind over matter. Think it into existence. If you believe it, you can achieve it. These sorts of sayings are more closely associated with self-help books found in thrift store bargain bins than they are with reality. But if you are willing to cheat a bit and throw some tech in the mix, you actually can make some really interesting things happen with the power of thought.

Take Thomas Vikström’s recent project, for instance. Using a Muse EEG headset and a machine learning algorithm that interprets its output, he demonstrated how a ping pong ball can be made to levitate just by thinking about it (and by turning on a blower motor with those thoughts).

To make this work, the Muse headband captures electrical activity from the brain across four channels, streaming the data into a Python-based pipeline. From there, the signals are processed using a spectral analysis algorithm from Edge Impulse, which transforms raw EEG data into frequency-based features suitable for machine learning.

The trained model classifies brain activity into three distinct mental states: calm, sleep, and non-calm. Each of these states corresponds to recognizable EEG patterns — such as increased alpha waves during relaxation or spikes in beta activity during cognitive effort or movement.

Once classified, the system maps these states to different power levels for a blower motor. A Particle Photon 2 receives this information via Wi-Fi and adjusts a MOSFET-driven blower accordingly. The levitating ball, stabilized by airflow and an assist from the Bernoulli principle, becomes a visible indicator of the user’s mental condition.

When the user of the system enters a calm, focused state, the ping pong ball rises steadily on a column of air. Drifting into distraction leads to fluctuations in airflow, causing the ball to drop or wobble. Relax too much, and the system interprets it as a “sleep” state, altering the behavior again.

This is more than a toy — the visible feedback can help people improve their focus. Research has long shown that mental fatigue and focus can be reflected in EEG patterns, and neurofeedback systems aim to help users consciously adjust those states. By turning invisible brain activity into something tangible, this project makes that feedback loop far more engaging.

If you’d like more details about the project, or would like to build your own copy of it, there is additional information in Vikström’s GitHub repository.