Single-Channel Brain-Computer Interfaces Sell Convenience, Not Intelligence

Last week I borrowed a new terminal from Wabo, and two of us on the team tried it for an afternoon. The most direct feeling is that it finally doesn't require conductive paste.

The most blocking part in this field is the preparation. EEG signals are easily drowned by physiological noise and environmental electromagnetic interference, so in the past, doing a collection involved applying paste, adjusting impedance, washing hair—pre-processing took at least half an hour, and the subject wanted to leave before even sitting still. Wabo made the single channel paste-free, which cuts out this segment. This is engineering work, but it determines whether this thing can leave the lab.

The rest of the account needs to be clear. Single-channel information is limited; based on my testing, it's more like a trigger—recognizing clear intent switches is okay, but don't expect continuous motion control. Don't be led astray by the words 'AI intent understanding'; the signal getting through is one thing, how much intent can be decoded is another.

To spread into rehabilitation departments and production lines, saving consumables and preparation time is more valuable than decoding two more intents. I've calculated the ROI: a technician running a few more rounds a day, and the equipment price difference is paid back in a few months. Next, watch the adaptation standards—Shanghai is already issuing clinical application recommendations, including personnel qualifications, environmental equipment, and operating procedures. Equipment is easy to buy, processes are hard to establish.

This step is in the right direction, but it only solves usability; there's still a distance to being good to use.