Today's work was focused almost entirely on the software side of the microscope, doing a lot of trial-and-error type troubleshooting of Python code and figuring out the best way to interface with the remaining parts of the Ender 3 Pro 3d printer from the attached Raspberry Pi. A fair amount of effort was also used to identify the parts of the microscope that will need special attention when designing future versions.
Because the microscope can be controlled in the Z axis, it is now possible to focus the microscope remotely. The curses library was used to get keyboard input through the VNC connection, and since the preview window is also visible on-screen, the operator can 'dial in' the focus and location of the bottom-left well without having to touch the microscope itself.
We were able to able to use the microscope to take photos of multiple locations within multiple wells, with a delay of only a few seconds between photos. This time can probably be reduced so that a full array of 49 photos per well on a 24-well plate can be done in less than an hour.
The lingering major issue is that our microscope's construction is extremely wobbly and fragile, which sometimes results in poor quality or simply incomprehensible output. The K'Nex prototype tended to bend out of place and cause the initial calibration to fail. It would also push around the even more fragile sample holder when trying to reach the 6th column or 4th row of wells, bumping the well plate from where the script expected it to be. We intend to address this by designing sturdier holders for the microscope and sample, to be assembled out of laser-cut acrylic sheets and acrylic cement.
A minor issue is that the printer's acceleration and movement speeds are jerky and tend to cause the samples and microscope components to wobble after moving. This can be adjusted manually on the printer's front panel, and can probably be adjusted remotely as well.