“Val, I could build one of these. A better one.”
Val’s colleague expertly placed the next mouse on the horizontal spindle of the Rotarod and pressed a button to start the run. The spindle rotated slowly at first, and then faster, and the mouse log-rolled to keep its balance on top of the rod. Eventually, the rotation was more than the mouse could handle, and it fell off the rod onto the chamber floor below. The researcher scooped the mouse up and put it back in the cage.
As we watched the experiment continue, I found myself seeing double. I was looking at the Rotarod the way an experimental biologist would: understanding how the device tested the animal’s coordination, what data it collected, whether the surfaces would be easy to clean, and whether it might frighten or injure the mouse. But at the same time, I was also starting to build the logical guts of the machine in my head: the software and hardware that would drive the spindle at different speeds and acceleration profiles, the sensors that would detect more details of the mouse’s behavior, the switches and buttons and screens the user would need. For the first time, I was seeing design in a piece of scientific equipment: looking beyond what a machine does, to how I might make it do what I want.
As I mentioned in my previous blog entry, I’m a research scientist who has gone into the bioinstrumentation business. If you’re interested in a similar transition, one of the cleverest things you can do is to cultivate this ability to switch between a user’s viewpoint and a designer’s.
You probably already have the user’s viewpoint down. If you have an advanced degree in biology, you have spent years training yourself to be a very close and logical observer. You also know, right down to your fingertips, what a well-functioning tool or machine feels like when you use it at the bench. The more unfamiliar part will be thinking like a designer: understanding at a deeper level what makes those machines work, and why they operate the way they do.
A good way to do this is to begin with a machine you already partially understand, even if you don’t realize it. If you’re reading this, the odds are good that you’ve used a micropipettor before. (The odds are good that you’ve used a micropipettor so much that your thumb joint ached for days.) And since you’ve been trained to observe, you know how it works. At the most basic level, a micropipettor is a syringe with a precisely adjustable spring-loaded plunger, so it’s not a conceptually difficult thing to understand — I’m sure you could give a sensible explanation of it to a seven-year-old. In fact, the first micropipettor was a modified syringe, and was created by a German biochemist in the 1950’s: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1369176/. (I encourage you to read this article: it’s a wonderful description of a clever and quirky scientist-inventor from the days when scientists more often made their own tools.)
When you started out in the lab, you were also trained how to use a micropipettor, and one of the things you were taught was to never, ever try to set the volume outside the stated range.
You may have been taught that it would break the pipettor, or that it would throw off the calibration, but I’m sure you were told not to do it. And from a user standpoint, that’s all you needed to know. But to gain a designer’s perspective, you need to look deeper. Why will the pipettor malfunction if you set it outside of range? To understand that, you need to look at the mechanical guts of the device.
Here’s some patent drawings of a mid-1970’s adjustable micropipettor.You can see that the device really is very much like a syringe with a spring-action plunger return. If you ignore the various gears and such for a moment, you can see that the top of the plunger shaft sits inside a hollow cylinder, and it’s pushed up into that cylinder by a spring wound around the central part of the plunger — sort of like the pen refill in a clickable ballpoint pen. Push the plunger down, and the shaft moves downward in the barrel and expels air, with the spring resisting the motion; let it go, and it springs back up, pulling air back into the pipettor shaft again. Easy.
OK, but this is an adjustable pipettor. It has to deliver a precise, repeatable aliquot of liquid over a wide range of volumes. How do you make that happen?
This is where we start paying attention to all of those gears. You probably noticed that the top of the plunger shaft is large and threaded like a screw (technically, it’s a worm gear), and that there’s a collar that’s threaded onto it. The collar is the plunger stop. It’s wider than the ends of the hollow cylinder the plunger is sitting in. When you take your thumb off the plunger, it can only spring back until the plunger stop hits the top of the cylinder.
The position of that plunger stop is what determines the volume. The hollow cylinder is turned with high precision when you adjust the volume, and vertical ridges inside the cylinder spin the plunger stop up and down on the worm gear, like turning a nut on a bolt. If the plunger stop is near the top of the worm gear (see Figure 12), it will keep the plunger from moving very much, and so the pipettor will only deliver a small sample volume. If it’s near the bottom, as it is in Figure 11, the plunger can move a much greater distance and will deliver a larger sample volume. You get precise results when you change volumes because the gears are well-meshed with each other and don’t slip.
But if you try to set the pipettor out of range, the piston stop will get to the end of the worm gear. Best case scenario is that the piston stop will spin in place, and the volume indicator dial won’t match the actual delivered volume anymore. (Pipettor recalibration time! Hope you didn’t have another use for that $100.) If you really try to force the adjustment wheel, the threads on the worm gear and the piston stop can also be damaged — a mechanic would call this “stripping” — which means that turning the thumb wheel won’t result in a smooth, accurate placement of the plunger shaft. The plunger will still move unless the damage is really bad, but you won’t know exactly how much it moved, so your accuracy is shot. If you’re very unlucky, the piston stop will warp and jam in the hollow cylinder, and you won’t be able to move the plunger at all. Now you know exactly what your bench training taught you to avoid.
This pipettor works perfectly well from a design perspective, especially if the user doesn’t abuse it. But as you look at it with your “user” hat on, I bet you can see a lot of design features that would make it unpleasant to use. (I’m not a fan of the thumb wheel for setting the volume, myself.) Subsequent pipettor designs have mostly focused either on making a pipettor more comfortable to use or making it cheaper, both of which are usability considerations. If you learn to jump between usability and design perspectives, you can home in on a truly useful device.
Oh, and the Rotarod? We tested my prototype a few weeks ago. It worked as I imagined it would, and the experimenters and the mice approved. Both had some usability suggestions, though, so I have more designing to do.