When scientists tracked great white sharks to a mysterious mid-Pacific hangout, they nicknamed it the White Shark Café — are they eating, or meeting? What’s going on out there?
Monterey Bay Aquarium (MBA) researcher Dr. Sal Jorgensen wanted eyes on the scene, so he and engineer Thomas Maughan at sister institution MBARI developed a Shark Café Camera tracking tag with funding from the Packard Foundation. Clamped to the dorsal fin, it detects diving and chasing behaviors, then triggers an off-the-shelf Sony Action Cam to record video, and finally pops off and floats for recovery. On its custom PCB you’ll find an ATSAMD21 microcontroller (same as Adafruit’s Feather M0 boards); a compass/IMU that senses location, acceleration, and even tail beats; an RGB light sensor; and a pressure-transducer depth sensor (Figure A). A capacitive sensor detects when it has returned to the surface, then a satellite modem phones home.
If you dare to tag a shark, you can build your own shark cam: it’s open source hardware and software. Maughan (a Make: reader!) taught intern Gabriel Santos a crash course in electronics he dubbed TechFest, then Santos built the camera tags and shared the project at github.com/thommaughan/sharkcafecam. Santos, now a biologist, continues to lead TechFests at CalPoly, teaching the next generation of biologists to use Arduinos and sensors.
Packard’s goal was to put better tech in the hands of biologists, so Maughan’s team shared their knowledge with Customized Animal Tracking Solutions to help them develop the CATS Cam (Figure B), now deployed by researchers worldwide on sharks, whales, manta rays, and sea turtles. Jorgensen’s work with CATS Cams revealed in 2019 that white sharks in South Africa don’t avoid kelp forests — myth busted — but cruise them routinely, bad news for seals and surfers.
Meanwhile, MBA’s California tagging program keeps producing amazing discoveries. Turns out even great whites are afraid of something: they’ll flee their feeding grounds for up to a year after a single encounter with an orca.
The shark tracker’s toolkit also includes archival tags (basic data loggers) and pop-up satellite archival tags (PSATs) which, like the camera tracker, release from the animal, float, and report their data to the Argos satellite network (clsamerica.com/science-with-argos). There are satellite positioning tags that report location constantly, GSM tags that report to cellular networks, and acoustic tags that ping a high-frequency unique ID code that’s picked up by receivers listening on the sea floor or afloat.
For do-it-yourselfers, David Mann’s Loggerhead Instruments shared their well-regarded OpenTag motion and depth logger at github.com/loggerhead-instruments/OpenTag3. And SparkFun now sells an open source Argos satellite transceiver shield that piggybacks onto their line of Thing Plus microcontrollers.
In recent years, scientists have increasingly turned to drone cameras to spot, follow, and record sharks, ranging from typical multirotors like the DJI Phantom to full custom underwater robots like Woods Hole’s REMUS AUV. But what’s really new is the addition of artificial intelligence to detect sharks automatically. The SharkEye project by UC Santa Barbara and San Diego State University uses aerial drone video and Salesforce AI to identify great white sharks at a beach with 95% accuracy and send notifications to the local community.