Resurrection

 "Do you think we can get it working again?" I asked Scott, as he opened up the Marsellus Wallace briefcase to reveal the bits of a disassembled brain-computer interface. 

Here we were on a Saturday morning, in Gary's stunning office space overlooking Saint Stephen's Green in the heart of Dublin, with the sun streaming through bay windows onto a desk just waiting to be covered with electronics gear.

A week ago, I'd visited the University College Dublin to meet with former colleague Ed Lalor, who, in the course of his Ph.D., is now refining the signal processing techniques we'd pioneered while members of the MindGames group at Dublin's MediaLabEurope.

At the lab, we'd designed and implemented a brain-computer control interface that could detect which region of the screen you were looking at. This game, Mind Balance, was our first public demo. When playing the game, a participant can control the character (Phil McDarby's adorable Scottish rogue, known affectionately as "Mawg") as he saunters across the tightrope wire. The participant does so using only their brainwaves. No keyboard, no mouse, no joystick. Just electrical signals measured non-invasively from the surface of their head above the occipital lobes.

Was it a parlor trick? As a game, perhaps. But imagine a world where your brainwaves could offer you another degree of freedom in a control system -- and think of how useful that freedom would be for someone who couldn’t use a conventional controller like a mouse.

So on that morning last week, Ed and I dusted off the old Symphony codebase that I cooked up back in the days when I still did cool things. Despite early signs of resistance, we scored a flawless victory, and got Mind Balance up and running. The Mawg sauntered across the tightrope wire, controlled by Ed's brainwaves, at a glorious and unprecedented 250 frames per second.

Which brings us to Saturday morning. Here we were, overlooking Stephen's Green, and planning to resurrect Cerebus, the wireless brain-computer interface which was once used to control Mind Balance. Scott had flown over for the weekend from his new gig in London, and together with fellow Mindgamer Ross O'Neill, and our former Principal Investigator Gary McDarby, we had every intention of giving it the good ol' demo-push try.

Ah yes -- The Demo Push. We haven't done one of those in a couple of years. But boy, do us MindGamers know how to execute the demo push. "Demo or die," was the old mantra, and do we ever have a motivation this time:

The Dublin launch party for Visual Studio 2005, SQL Server 2005, and BizTalk Server 2006.

It'll make more sense when I explain that Symphony's components -- including its signal processing, networking, graphics renderer, and Mind Balance content -- were all written end-to-end in .NET .

So with our mission objective clear -- the resurrection of Cerebus -- and with Ross jogging our memories by recalling the Pin code for the Cerebus Bluetooth module, we were off to the races. In no time flat, the Cerebus hardware rose from its slumber to launch into a happy chin-wag with the Symphony engine.   

But the signals coming off Cerebus looked pretty random, so we decided to take advantage of the hardware's versatility. Instead of measuring the electroencephalogram (EEG) brainwaves coming off of someone's head, which register on the order of microvolts, we decided to try acquiring a biometric that by comparison is transparently obvious: electrocardiogram, ECG, or in other words -- lub-dub -- a heartbeat.

We wired Scott up appropriately, and watched as his Iron Man heartrate came perodically thumping through to the screen. Among Scott Eaton's many superpowers, incidentally, is his ability to wield the strength of ten moose of above-average strength, who collectively are defending their turf during rutting season. But I digress, and yes, that ECG signal is caked in European-standard 50Hz noise. I'll add a filter into the signal processing network and sort that out in no time.

Fast-forward to Sunday afternoon, when Cerebus, seen here sitting on the now-cluttered table, was intermittently doing the business of measuring our heart rates. It was a tremendous victory. This prototype version of the Cerebus hardware was incomplete at the time that MediaLabEurope closed its doors. It is a real testament to the skill and tenacity of everyone involved that Cerebus worked this weekend at all - let alone as robustly as it did.

So -- all that remained was for us to (1) find a hapless victim, (2) fit Cerebus on their noggin, and (3) determine if our custom hardware would reliably monitor their brainwaves.

Like the Civil Engineer who's brave enough to be the first to trod across their newly-built bridge, Scott was first to don the device. Ross followed. 

There was a tense moment where we thought we'd erased Ross's early childhood memories, but it turned out to be a false alarm.

Suffice to say that our results weren't 100%, but we ended the weekend convinced that, given a few more days working at it, Cerebus could potentially be reliable enough to work for the launch event. And in any case, even if it's not connected to the main demo, we will set up Cerebus separately to wirelessly monitor one biometric or another.

In any case, Mind Balance is going to work at the launch event like a charm. 

So if you've ever wanted to try controlling a computer game character using only your mind, you're warmly invited to join us in Dublin on Thursday November the 10th (click to register for the Developer or IT Pro tracks). We plan to have Mind Balance running during the day at the Berkeley Court Hotel, and we'll certainly have it availble at the Ocean Bar (near the Grand Canal Dart stop) from 8:00 onwards. Even if you're not coming for the day's technical events, please don't hesitate to join us at the Ocean Bar and try it there.

In the meantime, I'll keep you posted. If it weren't for the million other things on our collective plates, and the fact we're all doing this in our free time, I'd see no reason why we couldn't get the Cerebus wireless interface up and running for the 10th.

But we have no excuse. Because by demo-push standards, two weeks of prep time is just short of eternity!

[update: here are photos from the launch event!]