Most ways of identifying a microorganism rely on how it looks. MicroMotion identifies it by how it moves — reconstructing the three-dimensional trajectory of a swimming organism and letting a machine-learning model read the pattern.
Under low magnification, two organisms can look almost identical — two blurred dots in the water. But they rarely move the same way. A swimming trajectory carries speed, turning, rhythm and three-dimensional shape: often enough to tell one organism from another with no fine morphological detail at all.
Because reading motion needs far less spatial resolution than reading an image, the approach opens a path toward low-cost, near-real-time classification — and, in principle, extends to any organism that moves and can be filmed.
Built during the project at the University of Granada, the lab records the three-dimensional motion of aquatic microorganisms non-invasively: high-speed video, a four-view mirror splitter for 3D reconstruction, a custom micrometric positioning stage, and fine control of light and temperature — everything needed to turn a swimming organism into a clean, trackable trajectory.
Every tool and dataset produced by the project is public and citable through a persistent identifier.