It's become much colder now in Heiligenstadt and the sun disappears at 4PM. Given that the institute is nestled in a valley between some rather high hills, I wouldn't say that it ever really gets to be full day light now: morning, noon, and evening all have the same appearance. Having always told the time by the sun, this is disconcerting to me.

The change happened suddenly, due to daylight savings time. The first few days, when I left work, I walked through the dark forest with my headlamp on. But the light seemed to emphasize the engimas of the trees and left me feeling a bit creeped out. So I did the logical thing, and started walking home without the light.

When I leave work now, be it early or late, it is always dark. And it is perpetually rainy, so neither the stars nor the moon grace the path with their presence. The mountains block out the town. If you have ever been in a cave when they turn out the lights, you have a good idea of what it's like for me to step out of my lab. But, if I let my eyes adjust long enough, there is a ribbon of off-colour black by which I can barely discern the path. When it is too dark in the woods to see even that, I guide myself by the grindy gravel beneath my feet. It's dark enough walking through the woods that my retinas play tricks on me and make colours that are not there dance before my eyes.

A few times I tried walking home along the highway. It is wider, so it is easier to follow and a little more of the non-existant skyglow filters down. But the road is narrow and the walks ended up being a little to adventurous as I had to run pell-mell alongside the road and then hop the railing to avoid cars.

Whereas I used to feel affirmation and joy looking at the nature around me while going both to and from work, now I can only see these things in the morning. But the mornings have become doubly beautiful for frost and my secret (don't tell anyone!) route to work brings me great joy. Enough, I think, to make up for it.

The institute has a load of new equipment, such as electron scanning microscopes, but also a bunch of legacy stuff. Occasionally, we find ourselves digging through odd drawers looking for Widget X and encounter some of these relics. Here my lab-mate sports a particularly SteamPunk-esque pair of glasses we found. Their true purpose remains unknown.

The institute derives a portion of its funding by tackling research problems for other companies. The problems are usually either too difficult or too experimental for the other company to spend time on, but still worthwhile enough to spend money on. My project is one of the experimental ones. Our goal is to determine whether a muscle's impedance (roughly speaking, its resistance to electrical flow) can be used as a separate control channel from the electrical signals which tell the muscle to contract.

I've put together a system which injects electrical signals into muscles and analyzes the response…

And tested this system on myself repeatedly, a practice which eventually led to my shaving all the hair of parts of my arms and my legs. By the time I was done, it may have been better to just shave everything, but then I would have been shaving in the name of vanity and not in the name of science.

And what I've found is that it is possible to use impedance to find out additional things about the muscle. Muscles can contract isometrically (a contraction without movement), concentrically (a contraction which shortens the muscle), or eccentrically (a contraction accompanied by a lengthing of the muscle, such as during slow push-ups or presses). All of these contractions are accompanied by noisy electrical signals as many hundreds of neurons simultaneously shout “Go! Go!” to their corresponding muscle fibers. This signal (in blue) looks the same for both isometric and concentric contractions, with the only major difference being the amplitude, which is dependent on effort and therefore not a good way to differentiate movements. However, the signal which I derived (in green), shows clear differences between the movements.

But the picture above is only one contraction. It's better to look at many and find a cunning way to combine the results. The image below is derived from fifty isometric and fifty concentric contractions, all of which have been normalized to the same amplitude and scaled to the same time interval. The resulting signal was then binned into a two-dimensional histogram.

The passive signal I extracted shows stark differences between the types of contractions, whereas the active signal does not.

So there's something there, and maybe it could be used to control a prosthetic, but more research, and more money, will be necessary first. We're hoping for the latter so we can do the former!

I feel like figures are one of the under-appreciated aspects of science. Its difficult to make them look the way you want and to convey your message. The figures above each took about three to four hours to make, including processing of the data. The data and photographs were taken in Matlab and moved to GNU Octave for post-processing and analysis. Axes, arrow, and borders were produced in LaTeX using TikZ. I started them during a train trip to Eindhoven and finished them at a very patient friend's house.

Check if this is a private message just for Richard: