Command the Box

One of the advantages of being old is that one can look back over time and see trends. But often, the time spent looking in the rear-view mirror comes at the expense of looking at what could be ahead in the road. Let’s start with a short historical tour of instrumentation. If someone put a random analytical instrument in front of you, you could reasonably tell – at least within a decade – when it was built, whether or not you were familiar with it. But not because of the instrument’s resolution or its sample size or even its detection limits. Rather, you would know because of the way you interacted with the system.

A hundred years ago, in order to conduct your analyses, you may have needed to place gratings in the right place, apply samples manually, add weights where appropriate, and so on. Then electronic instruments provided switches and knobs to operate. Likely a dial or meter appeared. Perhaps a light indicated the presence of a feature or condition. Indeed, buttons, knobs and dials were the operational interface for instruments for a while (and covered every available surface in fantasy star ships, such as the USS Enterprise, from the same era). Then came computer control and acquisition, so a keyboard (first defined keys, then “qwerty”) was affixed to the core instrument. Joysticks, mice, trackballs, and touch screens followed to provide what most would consider the modern instrument-human interface.

Now, think about every block diagram you have seen in a textbook. Over the last hundred years, the diagrams have become more graphically elegant and frequently more intricate as the instruments (and publishing technology) has evolved. But there is one very important piece of every instrument missing from those block diagrams: the user. Recall how you can tell when an instrument was made based on how you are expected to operate it? Perhaps more attention should be paid to the all important user-instrument interface.

Back to the present and future. What is next for human-instrument interaction? Initial attempts at eye movement, gesturing and voice recognition are already here for some cell phone users. I mean, what comes after that?

I propose that we begin using our cognitive processes and our ability to observe them as the method of controlling instrumentation. In other words, using our thoughts to control the box.

We have begun doing this in our labs and I see a wide expanse of opportunity ahead. There are several electroencephalography (EEG) headsets available. We have chosen to use the Emotiv EEG headset, which sells for a few hundred dollars – and even includes the software needed to create a pseudo-keyboard interface. You train the software by recording your 16-electrode EEG pattern when thinking a thought or command; when that thought pattern is duplicated, a flag is set that triggers whatever keyboard output you have previously entered. Pipe that keyboard output to the instrument’s software and – boom – you’re commanding the instrument to do a task simply by thinking the previously coupled thought.

I know what you’re thinking... But we have been able to make this very concept work on UV-Vis, IR, plasma atomic emission and NMR spectrometers, as well as on our GC-MS system. And we have gone well beyond simply telling an instrument to turn on and off. I have been able to teach students to take backgrounds or sample spectra, label peaks and conduct database searches all under the direction of only their cognitive processes. No muscular movement of any kind was required. Moreover, they all learned to do it within 15 minutes and without much problem (oh, to be a student sponge once again). So, what are you waiting for? Go tell that instrument to read your mind!