Considered some of the most vital organisms on Earth, Plankton microscopic are a class of typically microscopic plants (phytoplankton) and animals (zooplankton) which form the foundation of the global food chain. Beyond this, they also help fix between 30% and 50% of the world's carbon dioxide, making them critical in the fight against climate change. To study them, researchers often take small samples of water back to a lab where they are analyzed under a microscope to determine both the concentration and species of the organisms within. One group of people wanted to make this process faster and more accessible to amateur scientists, so they created the PlanktoScope project.
As the name implies, the PlanktoScope is a microscope for studying plankton samples suspended in a water solution. Two tubes sit at the top of the compact device, with one meant for the sample and the other to store what has already been seen. A pump sends samples from the reservoir into a transparent slide, after which a camera is focused on the illuminated sample and multiple photos are taken for additional processing and storage. The final step involves the pump moving the sample on the slide into the secondary reservoir and replacing it with a fresh one from the first.
To make this all happen, the team selected parts that would be both reliable and easy to access for nearly everyone. A Raspberry Pi 4 B acts as the controller for every other component, including a Pi Camera V2.1 module for taking photos, a stepper motor HAT, and a GPS HAT for receiving accurate timestamps and location data. Beyond the electronics and other miscellaneous cables/converters, a whole set of tubing, strainers, and connectors were purchased along with a slide for moving samples from one reservoir to the other.
At the start of the build process, a series of plywood components were cut from pieces of plywood and acrylic in order to form the shell and major internal structures. The slide assembly was mounted onto a set of two linear actuators, each driven by a single stepper motor that spins a threaded rod. Above sits the camera with its attached M12 25mm lens and a secondary M12 lens that can be changed to adjust the field of view. Finally, a 5mm white LED was fixed below the slide to cause enough light to pass through the sample.
Water is moved between the two reservoirs by using a peristaltic pump that is driven by a stepper motor. This style of pump was selected for two primary reasons, with the first being that its internals do not interact with the water directly through an impeller but rather by pushing the tube in a single direction. Second, the precise control afforded by the stepper motor allows for very minute adjustments to how much fluid is moving at any one time.
Once the sample is in place on the slide, the next step is to get an image of it and extract meaningful data for use elsewhere. The Node-RED application has several buttons that allow users to adjust everything from the focus of the sample by modifying how far away it is from the lens to the level of exposure compensation and even replacing the current sample with a new one. Once a new image has been taken with the Pi Camera module, the processing pipeline performs segmentation and edge detection to isolate regions of interest against the background. From here, images are saved locally and given the option of being uploaded to an external dataset.
To read more about this incredible project and all that its is capable of doing, you can visit the PlanktoScope team's website here.