Working under cover of night in parks as large as states, poachers are skilled in avoiding detection. If they kill with silencers on their rifles, the animal’s death is not likely to be noticed right away, even if it is wearing a tracking collar. Monitoring an animal’s tracking data is sometimes more of a check-in than continuous surveillance. Before rangers reach the location and confirm an animal has been poached, the poachers have fled the area.
This is the scenario that Vanderbilt University researcher Ákos Lédeczi and his team are trying to solve with an acoustic shockwave detection system.
“[Rangers] turn on the GPS only a few times a day,” Lédeczi says. “And they turn on the radio maybe once a day and they send the data maybe once a day.”
The system Lédeczi’s team is developing will always be “on,” so powering the device became the team’s biggest challenge.
“If you are using your phone, the microphone is constantly on and your battery lasts only a few hours,” Lédeczi says. “Now we have to listen to the sounds of nature continuously because we can’t miss the first shot—there may be only a single shot. This thing needs to last not for hours or days but one to two years on a single battery.”
For the past 15 years, Lédeczi has been working on acoustic gunshot localization. With the research wing of the United States Department of Defense (DARPA), his team developed anti-sniper technology that was not only able to report sniper location, but also to estimate and report the bullet trajectory, range, caliber, and weapon type with 95 percent accuracy or better. However, the battery of that system only lasted a day or less.
Lédeczi predicts the power solution they came up with for WIPER—a “sleep mode”-based concept—will last at least a year in the field on a single charge.
The Path to New Technology
Lédeczi got the idea for WIPER when he learned that elephant and rhino conservationists were looking into law enforcement gunshot detection systems for anti-poaching solutions. Systems such as ShotSpotter place an array of sensors on stationary structures—buildings, phone poles, lamp posts—and locate gunfire by triangulation. There are no buildings or phone poles in the wild.
“I came up with the idea, why don’t we put the sensors on the elephants and not the trees,” Lédeczi says. “And that way the sensor is always with the elephant.”
One of the conservationists looking for anti-poaching technologies was George Wittemyer, a member of the International Union for Conservation of Nature’s African elephant specialist group, associate professor at Colorado State, and chair of the scientific board at Save the Elephants.
Wittemyer had looked into the ShotSpotter system, but didn’t find it feasible for wildlife protection in large parks.
“The problem is, their system is really expensive, has a lot of hardware,” Wittemyer says. “We’re in really remote areas. The logistics of it are quite difficult—to set up and service it and to make sure that it doesn’t get tampered with.”
Wittemyer had even made several unsuccessful attempts to develop sensor technology for elephant tracking collars. So when he got the call from Lédeczi, offering to make such a sensor, he was elated.
What About Rhinos?
Though WIPER stands for Wireless Anti-Poaching for Elephants and Rhinos, the rhino anti-poaching device is a long way off.
“Collars for rhino have been problematic due to the nature of rhino skulls,” says rhino conservationist Jeff Muntifering. Funded by the Minnesota Zoo, Muntifering works with Namibia’s Kunene desert rhino, in a consulting role for Save the Rhino and the Namibian government.
“We have had some success using ankle bands with tracking devices here in Namibia, but I believe overall not successful enough to inspire further investment. I doubt those would be appropriate to attach an even larger device.”
“As far as I know,” Lédeczi says, “the only sensor they put on rhino is they drill a hole in the horn, and they put it in the horn. But for us, that’s too small. At this point, we need a bigger sensor, unfortunately.”
How WIPER Works
WIPER detects the shockwave of a supersonic bullet traveling through space, which cannot be muffled.
“Unlike existing military and law enforcement solutions that use arrays of microphones to accurately localize the shooter, WIPER employs a single microphone,” Lédeczi says. “The goal is to detect gunshots very reliably and provide an approximate location—in other words, that of the elephant being shot at.”
The device can detect shots from up to 50 meters away—roughly half a football field.
“Rangers can then be alerted and can cut off the escape route of the poachers to capture them,” Lédeczi said. “WIPER may not help the elephant wearing it, but it can prevent future poachings.”
The WIPER system operates with both GSM (cell) and Iridium (satellite) communication systems. It can send an alert along with approximate coordinates (depending on the most recent GPS location taken) within a second. Within a couple of minutes, it updates the location with the current coordinates.
“Unless,” Lédeczi says, “the poachers destroy the sensor in the meantime.”
The team will test the prototypes with Kenya-based Savannah Tracking, but ultimately Lédeczi and his team plan to release the designs and software into the public domain. The intention is to keep the cost down—to under $100—so that any tracking collar company can offer the WIPER tech as an option on its tag.
Testing Timeline
Over the summer of 2018, the WIPER prototype passed phase I testing on an elephant at the San Diego Zoo. There were no gun shots, but the team got to see how the device handled a wide variety of environmental sounds, including a lot of metallic clanging and elephant vocalizations, which can be sudden and loud.
“It was picking up sounds,” Lédeczi says. “Our initial shot detector algorithm didn’t quantify any of those sounds as a shot—so that’s a good start.”
Phase II, testing the device on one or two elephants in the wild, is slated to begin in early October of 2018 and will run about three months. Lédeczi hopes to retrieve the devices and have them back in the lab by the beginning of 2019, giving his team at Vanderbilt a full semester to sort through the retrieved data and make adjustments to the algorithms for accuracy. The main goal is to prevent false positives, which can make or break the tool’s success in the field.
“The reality with these types of technology in the field,” Wittemyer says, “the more false positives you get, the less credence users put into the device.”
The device has not been tested for ruggedness. It must stay on the elephant, take abuse, and still be able to detect sound, even through caked-on mud.
“Elephants are hard on equipment,” Wittemyer says, “strong, but also elephants continue immersing themselves in water, taking mud baths and dust baths.”
“So we are protecting [the sensor’s] core with gore-tex and steel mesh,” Lédeczi said. “These guys go through the thick brush and then the actual branches are hitting the sensor box—a loud sudden noise similar to a shockwave. Well, we’ll have to see whether that can survive in the wild.”
The device’s ruggedness will have a year in the field to prove itself. The yearlong trial of phase III is set to begin in early summer of 2019, so it will be at least 2020 before the product rolls out.
WIPER Deployed
“Sadly, humans are incredibly innovative at killing things,” Whittemyer says. He acknowledges the variety of poaching weapons elephants are up against, such as poison salt and harpoon traps.
“This system is not going to protect them from that,” he says. “But especially where there’s a lot of large-scale industrial poaching going on, that’s being done with assault rifles. And that’s where this system will be useful.”
There are a lot of ifs involved, but so far, WIPER is on track to help fight elephant poaching soon.
“I hope in three years from now,” Wittemyer says, “it’s like this is on every collar of every large mammal.”
This story originally appeared at the website of global conservation news service Mongabay.com. Get updates on their stories delivered to your inbox, or follow @Mongabay on Facebook, Instagram, or Twitter.
