I develop hardware and software for the researchers in the lab to perform experiments and for other related purposes. There are a number of different experimental arenas that I’ve developed, each containing some method to interact with a fly or flies. For example, we have a family of walking arenas that each use some combination of computer vision, robotic actuation, laser actuation, heat stimulus, visual stimulus, etc. These usually operate in closed-loop feedback, where the actuation and stimulus that the fly receives depends on what the fly does. The software platform that ties it all together is called Flylab, and while I won’t list the details here, it performs all of the tasks needed to run the walking arenas, including automating the experimental trials & data collection. I recently developed a very sensitive feeding sensor to distinguish instances when a fly is touching versus not touching food at a feeding station. Most of my time is spent programming in python, C, or Matlab, and I also provide other various lab services such as building miscellaneous electronic & mechanical instrumentation, occasionally working in the machine shop to make custom parts, fixing or developing the occasional bit of circuitry, etc. Ultimately though, it’s all just a lot of fun.
This video shows the realtime tracking and laser galvo system drawing numerals on each fly. In actual use we would use a different laser wavelength (e.g. 1064nm or 470nm) to activate the driver of the UAS-GAL4 system, and a small dot to focus more energy on the fly.
This video shows a fly walking on a circular arena with a “robot” maintaining a distance of 8mm. The robot is just a fly-sized magnet, and is actuated from underneath the arena via another magnet attached to a five bar linkage with two servo motors. This video is sped up 10x.