Mar 23, 2016Last week, I took a break from my day-to-day duties at IOT Journal to travel to Palau, a small island nation in Micronesia. It sounds like a vacation, but (aside from a lovely day of snorkeling) I was actually working on a freelance project: reporting for The Guardian on a research project that The Nature Conservancy is running in partnership with fishing company Luen Thai. The project's goal is to evaluate whether some small changes in how Luen Thai's tuna fishing boats operate might reduce incidental bycatch of non-target fish. (Bycatch refers to the hooking—or netting, based on the fishing process used—of species that are not the fisher's intended catch.)
But within hours of arriving, I realized that IoT technology is being employed as part of the research project. I guess my work follows me everywhere, even to remote islands in the Pacific.
Bycatch is a significant problem in fisheries around the world, not only because the fish that are unintentionally caught sometimes die—which, in turn, hurts the marine ecosystem, especially when bycatch include keystone species with low reproductive rates—but also because bycatch takes time and resources away from catching the target fish.
Sharks are often victims of bycatch by long-line tuna boats. When one is hooked, the crew cuts the line and releases the animal back to the sea (unless they are unscrupulous, in which case they might cut off its dorsal fin and sell it on the black market for shark fin soup). But what happens to bycaught creatures that are released intact? Does the trauma of being tethered to a fishing line underwater for some untold number of hours and then quickly yanked to the surface cause irreparable harm, or do the released animals tend to survive?
Research biologists have, for many years, used technology to try to answer that question, using various types of RFID tags and other approaches to track post-release mortality. But the most effective tools are called pop-up satellite archival tags (PSATs). These devices are designed to remain on an animal for a set number of days, weeks or months before automatically self-releasing and floating (popping up) to the water's surface, where they transmit stored data to scientists via a satellite link.
For the Nature Conservancy's research project, shark expert Michael Musyl, principal scientist of the Pelagic Research Group, is using PSATs to track the post-release mortality of two species—blue sharks and silky sharks—that are bycaught by the Luen Thai fishing vessel being used in the experiment. (PSATs are also utilized in general research, not just specifically to track bycatch mortality.) The tags being used, manufactured by Wildlife Computers, contain sensors that collect data regarding depth, temperature and speed of movement. A tag is set to release after 30 days, but if the shark significantly slows down for an extended period, or descends to a depth well beyond its normal range (in this case, 1,700 feet or lower) and remains there, the tag releases before hitting the 30-day mark. In those cases, the shark is presumed to have died.
It's a pretty handy technology, Musyl told me, but one that is in need of improvement. In fact, he and his colleagues conducted an analysis of 731 PSATs, manufactured by either Wilderness Computers or Microwave Telemetry, deployed in various marine research projects all around the world. They published their findings in 2011.
Data was received from only 79 percent of the 731 PSATs deployed. This indicates that in 21 percent of cases, the tags either failed for some unknown reason or perhaps functioned as they were supposed to but popped up in a place where the Argos satellite network does not provide adequate coverage, and researchers thus never received their transmissions.
But the really damning part of the analysis is that of the 79 percent of tags that sent data that was received, only 18 percent popped up after reaching their preset pop-up date. Does that mean that in the remaining 82 percent, the animal perished prior to the release date? Not even close. According to the data collected, only 2.3 percent of the tagged animals presumably died. That means that 97.7 percent of the tags popped up early due to some mechanical or electronic failure.
Musyl was quick to point out that failures are not always the manufacturer's fault. Sometimes, an infection emerges at the site where the tag is pierced into the target's body, causing the surface to erode to the point where the tag simply falls off. In other cases, some type of biofilm grows on the tag and compromises its structure, leading to a failure. He told me manufacturers are experimenting with coatings that contain antimicrobial agents designed to prevent this growth.
Unless the tag is physically recovered from the vast sea, which seldom happens, it's impossible to diagnose the failure.
Costs are falling, but PSATs are very expensive—between $1,000 and $4,000 apiece—and most marine research projects operate on tight budgets. Plus, when tags fail, the scientists' time and the resources required to find and tag the marine species are often wasted.
So my initial enthusiasm that advanced sensors were aiding important scientific and environmental research in a remote corner of the Pacific was met by disappointment that the technology is failing to meet researchers' desires.
Technology is always changing and—hopefully—improving. But hearing about the PSATs' performance limitations makes me wonder how they compare to wireless sensors used in industrial applications or other, business-focused sectors. What are the acceptable failure rates that your company or industry demands from wireless sensors or IoT systems? Send me an e-mail at mc@iotjournal.com, and let me know. Perhaps this is the kernel of an interesting investigation for IOT Journal.
Mary Catherine O'Connor is the editor of IOT Journal and a former staff reporter for RFID Journal. She also writes about technology, as it relates to business and the environment, for a range of consumer magazines and newspapers.
