Foraminifera–or foraminifera, for short, are protists, neither plants nor animals, and have been ubiquitous in our oceans for over 100 million years. When the stoma dies, it leaves behind its tiny shells, less than a millimeter across. These shells can be used to understand ancient ocean conditions, and the types and chemistry of holes in the sample are useful environmental indicators. However, separating different types of holes is a time consuming and tedious job that requires a lot of experience.
A team from North Carolina State University and the University of Colorado Boulder has developed and demonstrated a robot capable of sorting, processing and identifying microscopic marine fossils. New technology automates a tedious process that plays a key role in advancing our understanding of the world’s oceans and climate – both today and in prehistory.
The device, called Forabot, uses robots and Artificial intelligence To physically process the hole residue so that the residue can be isolated, photographed, and identified. The open source system can physically manipulate individual foraminifera for imaging and isolation with minimal human interaction.
Users first have to wash and sift a sample of hundreds of foraminifera, which results in a sample that looks like a mound of sand. The sample holes are then placed in a container called an isolation tower. A needle at the bottom of the isolation tower then exits through the sample, bearing a single puncture shell on its tip, where it is removed from the tower by suction. Suction pulls the puncture into a separate container called the imaging turret, which is equipped with a high-precision robotic device Camera which automatically takes multiple photos of the prospector. After images have been taken, the foramen is again lifted with a needle so that it can be captured by suction and deposited in the relevant container at the sorting station.
Currently, Forabot is capable of identifying six different hole types and processing up to 27 hole samples per hour. This may be slow, but unlike a person, a robot can do the task over very long periods of time without getting tired. They may be improved for classification purposes in the future by reducing image quality and/or quantity.
“This is a proof-of-concept prototype, so we’re going to expand the number of hole types it can recognize. We’re optimistic that we’ll also be able to improve the number of holes it can process per hour,” says Edgar Lobaton, co-author of a working paper on work.
At this point, this proof-of-concept prototype has an accuracy rate of 79% for identifying holes, which is better than most trained humans.
“Once the Forabot is improved, it will become a valuable piece of research equipment, allowing students ‘hole-builders’ to better spend their time learning more advanced skills,” Says Tom Marchetto, co-author of the paper and professor of geosciences at the University of Colorado, Boulder. “By using community-sourced taxonomic knowledge to train the bot, we can also improve the standardization of hole identification across research groups.”
- Turner Richmond, Jeremy Cole, Gabriela Dangler, Michael Daniel, Thomas Marchetto and Edgar Lobaton. Forabot: isolation and imaging of robotic wood foraminifera. Geochemistry, Geophysics, Geosystems, 2022; DOI: 10.1029 / 2022GC010689
#Robots #sort #marine #fossils #researchers #study #oceans #climate