CN Tower Light Show

During the 2010 Winter Olympics, the lights of the CN Tower are taking part in an interactive biometric art installation.

From 3,300km away, a participant in Vancouver has their brainwaves monitored as they watch a live video of the Toronto skyline. When they are inattentive, the CN Tower’s lights move at a lethargic pace. But as the participant concentrates, the lights of the CN Tower cycle around with the speed of Christine Nesbitt ’round the speed skating track.

The measurement being used to determine a participant’s attentiveness is their ratio of their alpha waves to beta waves. This very rough estimate of attentiveness is also what we used for an early project in the MindGames group called “BrainChild”, where a participant unlocked a door through sustained concentration.

What I like about the CN Tower installation (in addition to the cross-country nature of the biofeedback loop) was something I read in this Torontoist article. If participants are having trouble “concentrating,” they are encouraged to count the lights in skyscraper windows.

This is a very clever, indirect way to achieve the desired result. When working with “untrained” biofeedback participants, it can be difficult to articulate the subtle changes that they need to make which will lead to a desired signal (in this case, a change in the ratio of alpha to beta waves).

Mind Balance Training

Mind Balance Screenshot

A project we built at MIT MediaLabEurope called Mind Balance faced a similar challenge.

Mind Balance also relied on an electroencephalogram (EEG) metric to create a control mechanism. However, instead of measuring the ratio between alpha and beta waves, we were monitoring the occipital lobes at the back of the head to detect artifacts from the electrical signals produced by the brain’s visual processing.

The subtlety there was that success for a participant required not just having a visual pattern in your field of view, but also attending to that pattern – a “squishy” concept that required training. Some of the “indirect” methods we used to explain a “good stare” included encouraging a participant to “stare right through” a region on the screen.

We used a 45-second acoustic feedback training session to help participants learn this concept assisted by very clear and immediate feedback. It worked with varying degrees of success, but thankfully, enough success in over 95% of cases for a participant to successfully go on to generate a single reliable control axis.

We resurrected Mind Balance for the Microsoft Ireland Visual Studio 2005 launch and it was awesome (and somewhat improbable) that we got it to work in an Irish pub setting. You can read more about Mind Balance here, and check out the shots of our prototype headgear called Cerebus.

(Another MindGames project where indirectly achieving biometric results played a big role was Relax to Win. The techniques players used to achieve relaxation under pressure was an important and fun part of the interaction.)

Jedi Mind Tricks Re-Re-Rediscovered

Mindgames researcher Ed Lalor models Cerebus

Excitement around Brain-Computer Interfaces for commercial and artistic use seems to resurface every few years.

Although the headgear for the Olympic installation looks considerably refined compared to our prototype (pictured here), achieving reliability for novel and useful BCI metrics remains a really tough problem, principally because most people aren’t cool with having someone drill into their head, which is a shame, because the brain’s electrical signals are so much weaker by the time they reach the surface.

Later this year, with the imminent launch of Project Natal, the focus now seems to be on visual and acoustic interfaces.

But during the Olympics, it’s all about the Mind Games. My hat’s off to Interaxon for putting together this engaging cross-country installation.

(More about the Olympic mind-controlled art installation at PopSci, Interaxon, Torontoist)

(More about Mind Balance here with links to pics)