Eyes in the Sky

Bonus chapter from Poached: Inside the Dark World of Wildlife Trafficking

Reported September 2016

The plane hiccuped into a wobbly descent, bringing into focus the patchwork patterns of the burnt orange country below. Dirt roads stitched together a dusty landscape punctuated by dots of green mango trees and squares of brown thatch roofs. The scenery was beautiful in its vividness, but even from up here, the poverty was apparent.

The Ethiopian Airlines speaker system crackled into life, the voice garbled and indecipherable. A paranoid thought crossed my mind: was this really my destination? This did not look like a city of 1 million. Maybe this plane was making multiple stops, as is sometimes the case on less popular routes. I turned to the guy next to me—the one I’d been waging a silent, passive aggressive armrest war with over the last two hours, while he placidly read a book about Jesus—and cleared my throat.

“Excuse me, is this Blantyre?”

“Yes,” the man said, raising his eyes from his book. He looked me over, perhaps considering the implications of my question. “You have someone waiting for your here?”

“I do.”

“If not, you let me know. You’d be welcome in my home. Where are you from?”

“New York City.”

“You’ve come a long way,” he said. Our bodies heaved forward as the wheels touched down. I’d made it to Malawi. 

Sandwiched between Tanzania, Zambia and Mozambique, Pennsylvania-sized Malawi does not register on many outsiders’ radars. Indeed, prior to this trip I knew nothing about the place, and had never considered visiting. But I had been drawn here by a specific mission: to learn all I could about anti-poaching drones.

Around 2012, the media began hyping unmanned aerial vehicles (UAVs)—from little four-bladed quadcopters, of the type you’d get your geeky uncle for Christmas, to larger military-style gas and electric machines that look like Smurf-sized planes—as a “silver bullet” to end the killing of elephants and rhinos. The idea was to spot poachers from above and stop them before they could strike.

What’s usually not reported, however, is that the promising proclamations have failed to deliver. Conservation organizations big and small jumped on the drone bandwagon, but none have succeeded in catching even a single poacher using those would-be secret weapons. As a result, many former drone fans are disillusioned and now believe that UAVs have no role to play in the poaching war.

Today, just one group continues to seriously pursue drones as a tool for combatting poaching, a South Africa-based company called UAV and Drone Solutions, better known by its acronym, UDS. The UDS team is determined to prove the drone haters wrong—and they seem to be making headway. Of all the players in the field, they have by far the most experience, with 5,000 flight hours and counting. They’re also the sole organization deploying drones on a nightly basis, and with legal permission to fly the vehicles beyond line of sight.

“These are the only people we’ve found who are willing to really slog it out,” said George Powell, director of wildlife technology at WWF. “Day after day, they’re out there.”

UDS seemed the obvious choice for cutting through the hype and seeing for myself what drones can and cannot do, so I reached out to Otto Werdmuller Von Elgg, the company’s co-founder. A South African businessman of English descent who splits his time between Johannesburg and his family’s weekend retreat on the coast, Otto seemed an unlikely candidate to lead the conservation drone charge. But chat with him for even two minutes and his almost radical devotion to wildlife quickly comes across.

“I am an animal activist, not a conservationist,” he told me on our first call. “The distinction is I’m actually concerned about individual animals alive today. The more I get involved with the conservation crowd, the more I dislike them because most of it has very little to do with the animals, and instead with perpetuating a business. When I get miserable about this stuff, I just tell myself to pull it together because it’s about the animals, not the assholes involved with them.”

Besides animals, Otto ranks technology as his second great love, and wedding that interest with his passion for animals is what led him to wildlife-protecting drones. Otto and his co-founders began flying in South Africa back in 2013 and, aside from a 6-month period when the planes were grounded due to blanket government restrictions, UDS’ pilots have been flying pretty much nightly ever since.

In September 2015, South Africa became the first nation to adopt formal drone legislation and UDS, in turn, became the first company to gain permission under the new laws. A year later, they branched out to Zimbabwe—first group ever formally allowed to fly drones there—and to Malawi, where UDS is one of just two teams given a special exemption to fly. “We’ve done a lot of flying, and made a lot of mistakes, but we’re pig-headed and tenacious and will soldier on to make a difference,” Otto said. “This field is in its infancy, and we’re only beginning to understand how drones are going to be used effectively.”

Otto was keen for me to come check out one of the new field sites. Zimbabwe might be good, he said, because the country had gotten a lot of negative press recently over the killing of Cecil the Lion, a former feline celebrity at Hwange National Park. Cecil became an overnight internet sensation in 2015 when his controversial death was picked up by the media and went viral on Twitter and Facebook. The American trophy hunter who killed him—a dentist from Minnesota, who reportedly paid $50,000 for the privilege—bore the full brunt of the world’s indignant (and, many argue, misplaced) outrage. Telling Zimbabwe’s drone story, Otto said, would show that the country is in fact doing some things right in terms of conservation.

It was Malawi, though, that captured my attention. “Malawi is quite special, and quite a sad story,” Otto said. “Forty years ago it was the most beautiful country, but it’s essentially been deforested for charcoal, and now the national parks are these sort of jewels of all that remains.” The fact that I literally knew nothing about this place made me want to go there all the more, and I also figured this might be my only chance to do so. Malawi, then, it would be.

I quickly gathered that Malawi has had it rough for decades. The country fell under British control around the turn of the 20th century. The Brits dubbed their new territory Nyasaland (“Lakeland,” for the massive lake that dominates Malawi’s topography) and set to work implementing a host of racist policies, including giving 15 percent of the country’s land to European settlers and setting up a few parks for the benefit an exclusively white hunter clientele. When Malawians rose against the system in 1915, 40 people were executed and another 300 were imprisoned.

Independence finally came in 1964 under the leadership of a U.S. and U.K.-educated doctor named Hastings Banda. Though celebrated at first as nationalist hero, things quickly turned sour. Banda—who declared himself president for life—quickly checked the boxes for dictator, complete with all the repressive and totalitarian frills. He ruled with “a singular mixture of terror and ritualized paternalism,” wrote historian Maura Mitchell, and used religion to legitimize his actions. Torture and jail without trail were commonplace, and opponents were declared “food for crocodiles,” as the New York Times reported. By the late 1980s, one in five Malawian children died from malnutrition before his or her fifth birthday while Banda, meanwhile, amassed mansions, cars and helicopters.

In the early 1990s, Malawi was still a one-party holdout among a continent that was increasingly moving toward multiparty constitutions. But Malawians were becoming fed up. In 1992, seven Roman Catholic bishops banded together to speak out against the dictator, issuing a letter that called for freedom, government accountability and human dignity. Read from pulpits across the country and distributed as 16,000 copies, the Pastoral letter, as the statement is called, ignited a widespread opposition movement among Malawi’s citizens. Foreign donors, including the World Bank, further strengthened the pressure for change by pulling the plug on funds to the heavily aid-dependent country. Banda, somehow still operating under the delusion that he would win in an election, yielded to demands for a vote. In 1994, the long-time dictator was firmly defeated.

Even then, though, Malawi did not catch a break. The country cycled through various presidents, none of whom managed to significantly improve the economy, and civil unrest flared up from time to time (though war, notably, never broke out). Famines and floods periodically exacerbated an already difficult situation, at one point turning half the country’s landmass into a disaster zone. As such, Malawi remains one of the world’s poorest nations, with an average annual income is just $380. More than 10 percent of its 17 million citizens are HIV-positive and life expectancy hovers around a mere 58 years.

The reality of these statistics unfolded over the four-hour drive from the airport to the national park, from the young girl in the tattered pink dress who beelined toward me with outstretched hands, to the legless man at the gas station pumping air into the tires of his battered 1940s wheelchair. Countless pedestrians and bicyclists clogged the roads, reflecting the fact that Malawi has just 8 cars per 1,000 people. In rust-colored fields, men, the sun on their bent backs, scooped mud to make homemade bricks, while women sat on stools, selling bushels of bananas, tomatoes and red onions beneath a billboard proclaiming “Let’s Create a Progressive Nation!” above a headshot of the current president.

Not surprisingly, Malawi’s suffocating poverty has taken a toll on its environment. As Otto mentioned, the vast forests that once covered the country are almost gone, disappearing at a rate of about one percent per year to clear land for agriculture and to make charcoal, which people use for cooking and heating their homes. Indeed, many of those we passed on the road carried bushels of blackened wood strapped to their backs or balanced on rickety bikes—making their way to wherever they were going against a backdrop of barren hills deeply rutted by erosion. The country is trapped in a real-life prisoner’s dilemma, yet with each fallen tree, Malawi creeps one stump closer to a future marred by even more crippling poverty. How will people cook when the charcoal is completely gone?

Like the trees, wildlife has also fallen by the wayside. Long gone are the “splendid herds of the larger animals” that contributed to “the finest hunting country in the world,” as described by Scottish evangelical Henry Drummond in 1889. The losses began well before Malawi gained independence. Colonialists argued that “preservation of game was contrary to the agricultural development of the country” and—in the name of protecting crops—killed over 200,000 wild “vermin” in just over a decade, including elephants. As conservationist and lifelong Malawi resident G.D. Hayes wrote in the 1970s, “Any honest assessment of the reasons for the terrible destruction of Africa’s wildlife cannot do other than lay the blame squarely on the doorstep of the European immigrants.” In the case of any remaining elephants, poaching has more recently offered an unparalleled means of making quick money, not only for locals, but also for Zambian and Mozambican gangs that have infiltrated the country.

Wildlife is virtually extinct outside of the country’s reserves, but even within, animals are by no means secure. At Liwonde National Park, a protected area about the size of Chicago where I was headed to see the drones, poachers killed 50 elephants and two rhinos from 2014 to 2016. Countless other creatures were plundered for bushmeat. With almost no anti-poaching enforcement in place, hunters became ballsy and unafraid, illegally entering Liwonde during the light of day instead of under the protective cloak of darkness. They disassembled the boundary fence to make snare traps out of its wire and openly smoked the game they captured. Some even made themselves at home, setting up camp and staying for multiple days at a time as they amassed carcasses.

Luckily, the government acknowledged that outside help was needed if any of Liwonde’s animals were going to survive. In August 2015, Malawi enlisted the support of African Parks, the same non-profit organization that rehabilitated Zakouma in Chad (see Chapter Ten of Poached). Since taking over operations, the African Parks staff have pulled out some 18,000 illegal snares and 40 bear traps—one of which clamped onto an unfortunate ranger’s leg (he made a full recovery). Ten men have been arrested for ivory and rhino horn-related offenses, some of whom have been given jail sentences of five to 13 years, and over 100 others have been apprehended for infringements ranging from cutting trees to illegal fishing. One man, his court case pending, was a former park staff member who was found to be renting his gun to poachers. 

On top of the arrests, the African Parks team has also installed over 70 miles of new electric fencing around Liwonde’s border. This is meant not only to keep humans out but also animals—especially elephants—in. Around 2 million people live in the districts surrounding the park, and conflicts between villagers and wildlife are intensifying: in the last six years, crop-marauding elephants have claimed the lives of nearly 50 people. To further de-escalate the situation, in 2016 and 2017, African Parks arranged for over 350 of Liwonde’s 800 elephants to be captured and translocated to another park in central Malawi, one that had almost completely lost its elephants to poachers but is now also under the group’s watch.

With a year of runway and a handful of hard-won successes behind him, Craig Reid, Liwonde’s manager, now has the bandwidth to try more experimental methods of gaining greater control over the poaching situation. Solutions are especially needed at night, when the majority of poaching incidents now take place. Though things have improved over the last year, the killings and infringements have not ceased entirely. Liwonde has lost 24 elephants and one rhino since Reid and his African Parks colleagues took over—alarmingly high figures for a relatively small reserve.

“Yes, we do have too many elephants, but the situation we inherited was one where there was also an existing and growing illegal elephant poaching situation,” Reid said. Drones, he thought, could provide a way to preempt poaching incidents. “When you’re faced with as complex a situation as we have in this park, you have to look at all possible tools,” he said. “We really want to give drones a crack and see how they can help us.”

UDS was the obvious choice for undertaking that experiment. As Reid noted, “If you want to try something out, it makes sense to use people who know what they’re doing and have experience elsewhere.”

“I think we’re just more persistent,” Otto added. “We want to do the job.”

Dusk had cast the world in shades of lavender and pink when we finally reached a fortress-like stone outpost that marked the entrance to Liwonde. A ranger named Elizabeth—serious at first but then allowing herself a smile in response my enthusiasm—greeted us. She checked my passport and called Reid to verify my story before unlocking the iron gates and allowing us entry.

Crossing that threshold was like stumbling through a portal to the past. Outside, Malawi was all dust and trash, smoke and scrub. But here was the country as it had once been, wild and beautiful. I felt like Chihiro, the heroine from Spirited Away—all wonder-eyed and agape at the discovery of this new and seemingly impossible world.

I’d arrived at the height of the dry season, and hardly a speck of green could be seen. Autumnal piles of crackly brown leaves formed heaps beneath twisted Halloween branches; parched grey stream beds held not a drop to drink; and scant tufts of brittle, bleached grass poked up here and there like peroxide-fried hair. Night quickly overtook us and the driver slowed to a crawl, not so much to ensure that he didn’t stray from the worn dirt road, but rather to avoid hitting animals.

Yes, animals. Massive waterbucks, recognized for the distinctive white bullseye around their rumps, stared, deer-in-headlights-like, as we rumbled by, and bushbuck, their bodies spotted like fawns, scampered across the road. Whole families of warthogs trotted aside the car, tufted tails held up in attention. At one point, we stopped to observe what appeared to be a large, unruly ball of brown, grey and cream-colored feathers waddling through a field. In fact, it was a porcupine—the first one I’d ever seen in the wild.

Nearly an hour after reaching the park’s gates, we finally arrived at the guest headquarters, an airy, fenced off pavilion whose faint lights made only the most futile of dents in the unbroken darkness around us. A hand-carved board outside read “Beware of hippo and other animals.” A couple staff members with big smiles ushered me to my room, a five-minute walk away down a meandering sand path lit by tiny solar lamps of the sort you might find at a fancy upstate New York wedding.

Liwonde sees around 17,000 visitors per year, half of whom are international, and those guests generate 20 percent of the funds needed to operate the national park. Reid hopes to significantly increase their numbers in the coming years. More visitors will mean more jobs for local people, he pointed out, which in turn equals more money pumped into the surrounding villages. If the park is directly benefiting nearby communities, they will be more likely to do all they can to ensure Liwonde succeeds. A second but equally important slice of the tourist pie are Malawians themselves: a positive experience at the park creates a sense of national ownership, increasing the odds members of the public will care whether their wildlife lives or dies.   

After my long travel day, I was nearly as eager to finally meet Otto as I was for a drink. I found Otto waiting for me at the bar, in an elevated tree house-like space with wooden floors, pillow-covered chairs, walls decorated with animal skulls. He was dressed simply in a green t-shirt and khaki shorts, with grey hair cropped short.

 “Rachel. Can I get you a drink?” he said, looking up and smiling slyly.

This is going to be a productive trip, I thought.

Wine thankfully in hand, we headed downstairs where a candlelit table decked out in white linen had been set for dinner. All was dark around us, but in the morning I would find that that inky void contained an expansive view of a river floodplain, complete with hippos, crocodiles and warthogs. We were joined by a few tourists, including an older Canadian woman in the diamond business and her 20-something son and his girlfriend. At the end of the table, across from Otto, was another older woman from London—Chelsea, to be precise, she quickly informed us—who was bonkers about birds.

I was eager to begin grilling Otto about drones, so I was a little annoyed at having to make small talk with random holiday-goers. But the conversation proved entertaining, at least. Almost immediately, Otto launched into talk about wildlife issues, with all his characteristic candidness.

“In Zimbabwe, poachers are using cyanide to poison the wells, which condemns not only the elephants they’re after to a slow and excruciating death, but also all the animals that come to drink,” he told the aghast table over a plate of fish and vegetables. To amend the situation, he explained, he bought a handful of $300 cyanide-testing kits with his own money and gave them to the rangers. That wasn’t his first choice, though. “The solution I’d really like to do is to line the whole village up and then tell them, ‘Ok, everyone go drink,’” he said. “Those who refuse are the culprits.”

People at the table began getting into the conversation. “For the Chinese, we should just give them something that does work: Viagra!” the Canadian geologist said, referring to the widespread (and erroneous) belief that rhino horn is used as an aphrodisiac.

“I know Viagra works!” the English birder said, a bit too enthusiastically. Everyone laughed awkwardly.

As soon as the dessert plates were cleared, the tourists—who were getting up at 5 a.m. for a sunrise walk—went to bed. Otto and I lingered by the fire, chatting a bit more about the situation in Malawi while I finished my fourth glass of wine—much to Otto’s teasing, who turned out to be an exceptional enabler.

“It’s going to become more and more of a problem in Africa—these islands of nature pressured by starving people living outside,” Otto sighed, gesturing into the darkness. “Liwonde is a microcosm of what will happen in the future. I believe the best way to protect places like this are to have as few people working in the park as possible, because where you have people, you have temptation. People are the problem. That’s my take: the more technology you can replace people with, the better. I’ll be shot down in flames for that statement because they’ll say I have a vested interest and I want to sell drones, but that’s not it. The fewer conservation projects I’m involved in, the richer I am. This project costs me money; I don’t make money on it.”

“Why bother, then?”

“As I said, I’m an activist—just like my mum. I probably get it from her. She’s in her 90s and still going strong. She runs a vineyard on her own, has testified in court about all sorts of environmental issues and she don’t take shit from no one.

“Once, she saw a neighbor run over a dog and leave it for dead. But the dog wasn’t dead—it was whimpering, in all kinds of pain, but clearly couldn’t be saved. So my mum scooped it up and drove it to the neighbor’s house. She banged on this man’s door and said, ‘Hey! You ran this dog over. Now this poor animal is suffering. You must put it out of its misery. It’s your responsibility to kill it.’ But this big, tough-looking man, he couldn’t do it. So my mother said ‘Fine!’ She took the gun from him and shot the dog in the head—she didn’t hesitate.

“That’s my mum for you. She once told me she should have been born a man so she could get more done.”

“Well, it sounds like she really didn’t need a penis to make her way in the world!”

“No, she did not.”

We both gazed into the fire, silent for minute. “Well, you gotta rest up, girl,” Otto said at last. “Tomorrow you’ll see the drones.”

There’s a common misconception that drones are already patrolling skies across Africa, stopping bad guys in their tracks and leading to arrest after arrest. On my flight from London to Johannesburg, where I changed planes for Malawi, for example, the bespectacled South African gentleman next to me struck up a conversation. As soon as “drone” and “wildlife” came out of my mouth, he nodded as though he were an expert. “They’re regularly using drones in the Kruger National Park to stop poachers,” he told me matter-of-factly.

But this is simply not true. Drones are not even flying in Kruger anymore, let alone single-handedly stopping poachers there or anywhere else. There were drone trials in Kruger, yes, but management there largely considered them expensive failures. Nor have drones independently led to a single arrest in Africa; when flying blindly, in fact, they have spotted intruders just a handful of times out of thousands of flight hours.

The seeds of misinformation that sowed my seatmate’s drone convictions were planted when some conservation groups began publishing overzealous press releases tooting drones’ horns. Those messages were picked up by gullible media outlets, which spread the word far and wide, with headlines declaring “Drones Are Catching Africa’s Poachers” and “Rhino Poaching Stopped by Drones.”

UDS has indirectly gotten caught up in some of the hype, by way of one of its funders, Air Shepherd, a program of the non-profit U.S.-based Lindbergh Foundation. One of the first reserves that UDS flew in with Air Shepherd’s support had previously suffered up to 19 rhino poaching deaths per month, but in the six-month trial period that the drones flew, no rhinos were lost. Although encouraging, Air Shepherd over-interpreted the significance of those results. “Our track record is 100 percent,” said Tina Pirazzi, the group’s director. “We know it works. We’re making a very real and measurable difference in the lives of animals. It’s extraordinary, really, a wonderful solution to a tragic problem that needs to be shut down.” The group’s website equally exaggerates the successes, triumphantly declaring, “We use drones that can fly at night and find poachers before they kill. It works. It’s proven. The poaching stops.”

While Pirazzi and the Air Shepherd team’s hearts are no doubt in the right place, there is “no real statistical proof for the usefulness of drones for anti-poaching—zero,” according to Nir Tenenbaum, director of Wildeas, a conservation-technology consultancy. Even if the area where their drones patrolled was cleared of poaching for a time, that fact unto itself does not prove causation. Without controlled trials, those observations remain firmly in the realm of correlation. There’s also a second, more fundamental problem with Air Shepherd’s and others’ inflated statements, Tenenbaum added: there is no such thing as 100 percent in security. Claims to the contrary are actually impossible.

Despite Air Shepherd’s tendency toward exaggeration, none of the outside experts I spoke with questioned its creators’ dedication to saving wildlife. That’s not always the case in the drone industry, though. When anti-poaching drones began gaining traction a few years ago, numerous other groups jumped on board, including technology businesses eager to push their products off on naive conservationists. “The salesmen from many of these companies trying to sell UAVs are pretty good at talking people into things, but then they don’t work,” said Powell at WWF. Many projects barely got off the ground before being declared failures.

Ol Pejeta Conservancy in Kenya, for example, launched a UAV trial with the help of a U.S. drone company, only to find that the planes easily broke down in the rugged African terrain. The camera they carried wasn’t good enough to detect poachers and they couldn’t fly long enough to be of much use. “We didn’t feel that drones added very much to our overall security capability,” said Richard Vigne, Ol Pejeta’s CEO. “I don’t think the world of conservation has the money to spend that would be needed to make an anti-poaching drone effective.”

WWF was also an early player in the drone game. After winning a $5 million grant from Google—an award, in typical Silicon Valley parlance, meant “to support organizations using technology and innovative approaches to address some of the toughest human challenges”—WWF reps traveled to Namibia to carry out trials using Falcon UAVs. From the beginning, though, Namibia was “absolutely a disaster,” said a drone expert with knowledge of the project, who asked not to be named because he was not authorized to speak on behalf of his employer.

In hindsight, WWF also gingerly admits things did not go as planned. “There was a lot of hyping of drones in the military context of using them in Iraq and Afghanistan, and people weren’t sure of what they were getting themselves into,” said Colby Loucks, lead of WWF’s wildlife crime technology project. “Drones were so new at the time that people had false visions of what they could do. They just heard the word ‘drone’ and thought they were much fancier than the civilian-grade ones we were using.”

The Namibia project was soon scrapped, the official reason being because the government suspended the use of drones. Some, though, see Namibia and other UAV failures as indicative of a universal problem that regularly stymies engineering undertakings of all sorts: a tendency to choose solutions first and then try to force a success, rather than analyze the problem and then use the findings to select appropriate tools. “So many groups want technology to solve all their problems, to save the rhinos and elephants,” Tenenbaum said. “Usually, though, they don’t understand the technology, and they don’t take the time to build a plan with someone who knows what they’re doing.” In other words, wildlife managers fall in love with the idea of using drones without really knowing anything about drones.

Funders also drive interest in trying to force a UAV solution. A friend of mine at a wildlife non-profit, for example, said she regularly fields calls and emails from Western donors who read a rose-colored 400-word news story about UAVs and then suddenly want their money to be redirected toward that perceived solution. Conversely, no one wants to support “an analytical process,” as Tenenbaum calls it—even the term sounds boring. “But if you don’t go at it with the right process, then you will lose the game,” he said. “It’s very frustrating, seeing again and again and again how organizations just throw away millions of dollars—millions—because they decide beforehand what the solution will be.”

After so many failures, a significant number of conservationists in Africa are no longer interested in dealing with drones at all. But this doesn’t mean the technology has no role to play in conservation. Indeed, a period of disappointment is actually standard for most new technologies. As the U.S. research company Gartner once famously observed, new innovations tend to follow a clear hype cycle that takes a realty-check-driven dip before the tech either finds its place in the world or flops. Drone technology itself has also become more sophisticated and affordable in a few short years, increasing the odds that the machines will find a useful role in combatting poaching. As Loucks summarized: “There’s been a lot of arm-waving about drones being awesome and that they’ll save the world, which turned out to be wrong. Then it was drones stink, let’s scrap them all. But I think there’s a happy medium.”

 “I think the original vision was correct and the Google investment was wise, but you have to fail fast,” added Crawford Allan, senior director of TRAFFIC at WWF. “You go in there, learn, move on, change and adapt.”

Fail fast: that’s a lesson Otto has gotten to know well over the past four years. Although UDS has managed to stay in the game longer than anyone else, he and his pilots’ time in the field has been steeped in frustration and disappointment, with a few precious successes thrown in.

From the beginning, there were technical hitches. Otto and his partners suffered a $150,000 whoops when they purchased two Penguin Bs, petrol-powered military-type drones. The fancy machines turned out to be completely useless in the bush. They couldn’t take off or land on an improvised runway, they were noisy and their constant need for fuel quickly added up in terms of expense and logistical hassle. Unfortunately, Otto and his partners couldn’t just stick the drones on eBay, so recouping their money wasn’t really an option. The Penguin Bs are now in Otto’s office, serving as exorbitantly priced coffee tables.

Learning from that mistake, UDS turned next to BatHawk drones. Smaller, hardier, quieter and nimbler, the planes proved much more appropriate for the rugged conditions that the team needed to operate in. They’re also relatively affordable, at $15,000 a pop. Rob Hannaford, Otto’s tech-savvy business partner, further strengthened the drones’ design with bespoke modular components and created surpluses of 3D printed backup parts. “A lot of companies want to sell very pretty, very sexy drones, but ours is not,” Otto said. “It’s built to work in the field.”

Kruger National Park served as the team’s first testing grounds. But excitement soon turned to disappointment—and even resentment. The group’s time there was fraught with challenges, and they quickly learned that their preconceptions about swooping in and solving the poaching crisis were woefully misinformed. Though Kruger is ground-zero for rhino poaching in Africa, it is absolutely massive—about the size of New Jersey or Wales. Flying blind, the drone pilots spotted poachers about one in every 500 hours of flight time—a hopeless and pointless undertaking. “Don’t give us a place the size of Israel and say, ‘Go find poachers,’” Otto said. “It’s just a waste of time. It’s turning batteries into noise.”

Patrolling without intelligence clearly didn’t work, but UDS was dismayed to find that management seemed wholly uninterested in working with them. The head of rangers at the park never took the time to join the UDS pilots in the field to see what the drones could do, and neither did management allow the drone operators to conduct joint missions with Kruger’s anti-poaching personnel. Yet UDS also was not allowed to operate on their own; they always had a handler with them, directing them where to go. One of those handlers was an award-winning senior ranger named Rodney Landela—a “poster boy for rhino anti-poaching,” as Otto said. In July 2016, however, Landela was arrested after he was caught allegedly poaching with one of the veterinarian technicians.

Once—just once—UDS pilots did randomly spot suspected poachers from the sky. But when they radioed in the finding, they were told that no ranger teams were available. For the drone pilots, the experience was very demotivating. As Antoinette Dudley, one of the pilots working that night, later said, “At Kruger, searching for poachers was like looking for a needle in a haystack. So if you find that needle, it’s a big deal. But then when no one responds you realize all that time has been wasted. It’s a bit of a downer, it’s not a nice feeling. You get upset with people because they say they want you there, but they don’t back you up.”

Those working at the park saw things differently, however. Colonel Otch Otto, Kruger’s former mission area operations manager of ranger services, did not mince his words when asked about the incident:

“They didn’t see anything for seven months, so I didn’t allocate a team to them. I’m not in research, I’m in operations, and dedicating a reaction capability to an unproven technology does not serve the rhino or the operational objective of the day. The remark that no teams were available to follow up is arrogant. Otto is very enthusiastic about his equipment, but drones cannot do the things TVs told us that drones could do. I’m of the opinion drones have limited applications and they’re expensive. They’re also unproven and inefficient—and I don’t have time for that. UDS will not save a single rhino with a UAV for the next two years, I guarantee it. Bring on the drones when they are a leading game changer and a competitive asset. But until then, I don’t want a drone on my land. If we have extra money we will buy a fence, not a toy airplane!”

Regardless of what really happened in Kruger, it is fair to say that the park’s size—both literal and figurative, in terms of its many layers of bureaucracy—likely contributed to the poor results. “The more you dig into Kruger, the more you understand that it’s wheels within wheels,” Otto said. “It’s just a quite toxic environment. People who are supposed to be protecting the rhino are not.” In May 2016, funding for the drone work at Kruger dried up and UDS pulled out.

Otto sees Liwonde as a sort of antidote to the Kruger situation: the park is small, managerial support for drones is strong and the pilots are allowed to operate without direct supervision. Importantly, Liwonde also has no after-dark monitoring capabilities. If UDS can’t get drones to work in this most ideal of petri dishes, they probably won’t work anywhere. 

“Drones are clearly not the silver bullet to end poaching, but we are very confident we’re onto something,” Otto said. “Anyone who says drones aren’t at all useful is being a clot, because we’re giving them capacity to operate after dark. Nighttime is the big boogeyman for poaching.” 

For now, though, that’s a gut feeling based on anecdotal evidence; Otto concedes that the lack of data to back his assertion remains a problem. To amend that, WWF and UDS are planning experiments in Malawi and Zimbabwe using some of the Google grant funds. The plan is to test drones’ usefulness in scenarios ranging from illegal fishing and poaching to human-elephant conflicts. Powell is clear about the agenda: “We went into this with the understanding that we’ll produce scientifically rigorous reports and make them public—that is the deal. Hopefully we’ll save some elephants in the meantime, but we’re going to do this as scientists.”                                                                                                    

Weeding out correlation from causation will be high on the priority list. If poaching declines in an area where drones are flying, the knee-jerk reaction is to say that drones caused this welcome turn of events. But that conclusion is premature; even if poaching drops or stops where drones fly, that does not mean drones actually drove the change, or that they impacted poaching overall. Animals may no longer be present in the places where drones are flying—they’ve moved or have been poached out of existence, so deaths in that area become zero regardless. Or it could be that weather, economic changes, increased law enforcement or numerous other factors that have nothing to do with drones caused poachers to put down their guns.

The most likely scenario, though, is that drones create a deterrence effect that drives criminals to different areas where they know they’re not being watched from above. Wildlife managers refer to this phenomenon as a “diversion of poaching.” If this is the case with drones, it means that those expensive machines are just moving poachers around, rather than stopping it altogether. Indeed, when one notorious poacher was finally apprehended by rangers in a KwaZulu-Natal reserve where UDS formerly patrolled, he confirmed that he had known when and approximately where the drones would be flying, and strategized accordingly. When UDS ended its trials, he resumed his activities. Poachers are not dumb.

On the other hand, if it is possible for drones to create a net reduction in poaching, the researchers want to know exactly which ingredients contribute. In how large of an area do drones effectively operate? In which terrain do they work best? What gear should they be equipped with? What type of public relations campaigns at local villages work best? The list goes on. Rigorous scientific trials can begin to address these questions and provide a road map for moving forward. As Powell said, “At the end of the experiment, we want to be able to say, ‘Here’s some good things you can do with drones and here’s the things you shouldn’t try because they probably won’t work.’”

At Liwonde, the time had finally come to investigate those answers.

Stephan De Necker and Antoinette Dudley are bush people. For those (like me) who are unclear on that term, they’re the type who hate cities and relish the wilderness. For them, the rougher the living, the better. No electricity? They’ll build a fire and heat their canned beans over the flame. No bed? The ground’ll do fine. No toilet? Ha, they’ll dig a hole! No other people? Perfect, they only need each other. As Antoinette said, “We’re together 24/7, and we’re actually quite chill. The only thing we ever fight about is if there’s no hot water and Stephan can’t manage it.”

I met Stephan and Antoinette, two of UDS’ senior drone pilots, late in the afternoon for a chat—up all night working, they had been sleeping until then. They’d arrived at Liwonde just three weeks before, lured by the promise of travel and adventure. A journey that had consisted of a quick two-hour flight from Johannesburg to Blantyre for me was a “nightmare” for them, involving 12-hours per day behind the wheel for four days straight, including three international crossings. Though they strategically concealed their drone gear under blankets and pillows to avoid questions from suspicious officials, they still wound up doling out bribes to several unscrupulous characters.

“You learn to bypass certain things like the border control,” Antoinette casually explained, as though she were talking about sneaking into the express checkout line at Whole Foods. Along the way, there was just one close call, when a man at the Malawi border noticed a piece of drone gear. “He looked at it and said, ‘What’s that?’” Antoinette recalled. “I just said, ‘That’s our stuff.’ There were no more questions after that. Offering him a cold drink also helped.”

“Sometimes you need to do it the African way and pay a bribe,” Stephan agreed. “Just give the guy $50 and drive.”

“$50! Stephan, it was $160!” Antoinette corrected him.

The two of them make a cute couple. Antoinette is slim and petite, with dark eyes and long brown hair that she ties back in a loose pony tail. Her features reminded me of Audrey Tautou, though less delicate—the result of years spent in the sun and hard work in the field.

Stephan had handsome features, like a stockier Liev Schreiber. Puffs of snowy chest hair peaked out of his green button-up work shirt, which featured a UDS patch on the front. When I asked him why he likes to spend time in the bush, he shrugged. “I was probably born like that. I don’t know any other way to express it. You love it or you don’t.”

Stephan is not a big talker, and when he does speak, his accent gives away that English is not his first language. As Antoinette described him, he is “a proper Dutchman”—though she added that that can be a derogatory observation, depending on how it’s expressed. “When you’re English and you speak Afrikaans and make a mistake, you sound cute,” she explained. “But when you’re Afrikaans and you make a mistake in English, you sound dof.”

“What’s dof?” I asked.

“That means stupid,” Otto interjected helpfully. 

Otto worked with Stephan years ago and remembered his exceptional skill with machines, as well as his love of the bush. So when the drone pilot positions came up, Stephan was one of the first people Otto called. As for Stephan and Antoinette, the two of them met through a mutual friend from the South African police department (they both have police backgrounds). Seven years later, they’re still going strong. 

Couple teams tend to work well, Otto has found, ensuring against loneliness in the field. It’s an added bonus if they are nomads with no children and no fixed address. Otto has also learned the hard way to recruit tech savvy wilderness lovers who then learn to become pilots, rather than the other way around. “We tried to take pilots and turn them into bush people and we had guys sobbing after two months, saying they couldn’t take it anymore,” Otto said.

“You get these guys from the office who are great with planes and you put them in the bush and they just crap,” Stephan agreed. “They say, ‘Shit’s not for me! I’m leaving.’”

“‘I want McDonald’s!’” Antoinette laughed, miming a whiney voice.

Passion is the job’s final prerequisite. Without it, pilots simply do not find the rigors worthwhile. To keep poachers on their toes, the hours are erratic, with no set schedule: they may start work at 6 p.m. one evening, 3 p.m. the next day, and midnight the night after that. “The biggest problem is sleep,” Antoinette said. “We go to bed at 5 or 6 a.m., but by then it’s already hot. There’s no AC, no fan, no electricity. We sleep under a wet blanket, then when the water evaporates we get up and wet it again to keep cool. You do toughen up a bit when you’re out here.”

“TTFU, as I say,” Otto chimed in. “Toughen the fuck up!”

Antoinette laughed. “Stephan wants to feed me a teaspoon of cement every day.”

She and Stephan previously piloted drones at Kruger National Park, and then in KwaZulu-Natal, UDS’ second South African site. While rhino killings in Kruger dipped slightly in 2016, in KwaZulu-Natal, the killings are on the rise. “In Natal, a few years ago, there were only a few rhinos poached a year—they could handle it,” Antoinette said. “But in under a year, it became a crisis. The same is coming with elephants, giraffes and lions. Because the Ching-Chongs can’t eat normal fruits and veggies.”

“I want to kill Chinese people and make them eat each other,” Otto muttered. I wasn’t sure he was joking.

Political incorrectness aside, that Antoinette and Stephan harbor a fierce desire to protect their continent’s wildlife is unquestionable. By the time I arrived, they’d already intersected a few illegal fishing boats and discovered that small, multi-rotor DJI Phantom 3 drones—the kind you might see flying around a city park—could act as electronic border collies for elephants, herding would-be crop raiders away from fences. Elephants hate bees, so it could be that the whirling buzz of the Phantom’s quadruple blades makes elephants think their six-legged nemeses are around.

As we chatted, the sun dipped low over the floodplain, once again casting Malawi in the soft pinks and purples I’d seen on my drive in. It was time to head to the field.

Stephan and Antoinette’s Land Cruiser doubles as the couple’s cockpit, and it was impossibly tricked out with meticulously organized wires and gear, utilizing all surface space. Tools, plugs, safety goggles and even a mosquito zapper were neatly held in place on to the car’s plush ceiling with Velcro, while the trunk contained perfectly wound coils of variously colored wire, rows of batteries, a sun visor and drone parts. The set-up would make Marie Kondo proud. 

We rumbled down a bumpy dirt road to a riverside field. Stephan and Antoinette quickly checked the area for animals and then launched into action. A folding metal table became an outdoor workbench upon which they expertly assembled the BatHawk, working around each other the the perfect unison of tango dancers. Front propeller, gimbal camera holster, video transmitter, battery, telemetry device and radio-controlled antenna were all attached to what looked like a four-foot long silver and black missile. The wings came last—the ah-ha addition that finally revealed that this was indeed a plane. All told, the drone weighed just under 9 pounds.

Next, Stephan climbed onto the car’s roof and hooked up a thick 8-foot tall antenna while Antoinette got to work on the computers inside. Two monitors, one strapped to the back of the driver’s seat, the other the passenger’s, provide them with a pilot and video display. The BatHawks fly on an automated, pre-assigned flight path controlled through an open-source program called PixHawk that Rob, Otto’s techy partner, further refined for UDS’ purpose. Once Stephan manually gets the drone into the air he doesn’t have to manipulate its course again until landing, unless he and Antoinette spot something of interest and want to veer off course. In this way, the job resembles less that of an F-16 fighter pilot, more that of a commercial airline pilot who puts his legs up after takeoff.

The Lisa Frank sunset melted into velvety purple, signaling time for takeoff. Stephan and Antoinette performed a series of pre-flight checks, just as any pilot would do. Satisfied that all was in order, they prepared for the launch. This was a quick affair, which Stephan made look effortless: grabbing the belly of the plane, he lifted it up, its engine quietly whining, and then shot it into the air like a javelin. That was it. The BatHawk zipped away, higher and quieter with each passing second, and finally disappeared entirely.

The two pilots hustled to the car and took their seats in the back. Stephan’s monitor displayed a series of yellow lines overlaid on a black and white GPS image of the terrain—the plane’s flight path. Antoinette’s showed a monochrome live stream view of the earth below, one visualized in thermal signatures rather than color. Warm things—a human body, for example—showed up as black, cooler things as white. Viewed this way, the river appeared light grey, trees dark grey and the land chalky white.

“Are you controlling it now?” I asked Antoinette, curious about the old Playstation controller in her hands.

“No, this controls the gimbal holding the camera,” she said. 

“We didn’t show you the little Chinaman who’s sitting inside the plane controlling it!” Stephan cracked, keeping his eyes on the screen. “No—the plane’s flying by itself.”

These skills did not come easily. As with all 16 of UDS’ pilots, Stephan and Antoinette had to undergo a huge amount of upfront training and investment. First, they acquired a private pilot’s license—the same as if they intended to operate manned commercial aircrafts. That required 40 hours of flight time plus ground school. But while it’s possible to go from zero to private pilot in about two-and-a-half months, UDS’ pilots usually take six to nine months before they’re ready to deploy to the field. After acquiring their private license, they must undertake additional internal training to learn how to operate the drones, not only on a nice, flat tarmac on a beautiful sunny day, but also at night and in difficult conditions. At the same time, they become versed in the software and in interpreting the imagery from the live feed: they must be able to distinguish a tent from a rock, an elephant from a hippo, a human from anything else the thermal camera might pick up. Finally, they need to be familiar enough with all the ins and outs of the tech to perform a fix when something inevitably breaks in the field. As Stephan said, “It’s not if you’re gonna crash, it’s when you’re gonna crash.”

UDS is licensed to fly the BatHawk up to 15.5 miles away, but they usually stick to about 9 miles maximum, just as a precaution. Batteries last up to two hours, and with rechargeable replacements, they’re able to fly more than 8 hours in one sitting. While the planes can book it at 50 miles per hour, the pilots prefer to take things “low and slow,” Stephan said, flying at about 30 miles per hour and just high enough to be out of earshot.

The two were explaining all of this to me when suddenly, an odd form appeared on Antoinette’s screen. Popping out darkly against the river’s luminous grey was a pod-shaped form containing two black bumps. A boat! With people in it!

“They’re breaking the law by coming into the park,” Antoinette said, leaning forward to scrutinize the screen.

“Let’s give them a scare,” Stephan said. With the tap of a few keys he broke out of autopilot and sent the drone circling back toward the boat, descending from its 300-foot cruising altitude. He switched on the navigation lights, lifting the BatHawk’s cloak of invisibility against the dark sky. 

The reaction was instantaneous: the boat did a U-turn, sending it back toward a fishing village on the park’s border. Like the slime trail a slug leaves behind, in their wake the intruders left a ribbon of thermally disturbed, slightly darker grey water. Though the men were not apprehended, perhaps they’d think twice about illegally entering the park again in the future. “That’s the takeaway after three years,” Otto said. “Drones are something in the night sky that wasn’t there before.”

“To make a difference, you’ve gotta be visible,” Antoinette agreed. “They’ve gotta know you’re there.”

Prior to witnessing this demo, some conservationists I interviewed questioned the ability of thermal cameras to produce footage that is good enough to distinguish humans from animals or landscape features. At least in the trials I saw that evening, the difference was obvious. The men in the boat were clearly visible, as were our forms when Stephan flew the drone over the field where we were standing. Animals differed in shape and size, as well as in the way they moved. Human intruders, Stephan added, also tend to register as a deeper shade of black, indicating a higher body temperature that reflects the fact that they are hustling to get in and out. With practice comes the ability to recognize these differences and pick up on anything out of the ordinary. “Even a dog is suspect, because they shouldn’t be there,” Antoinette said. “Meat poachers use dogs.”

Despite the pilots’ skills, image analysis remains problematic because it must be done in real time. In other words, if a drone flies for eight hours, its handlers must watch eight hours of continuous video feed. Look away for a second and they might miss an intruder. Such was recently the case in KwaZulu-Natal, where rangers discovered evidence of an attempted poaching incident that had taken place the previous night in an area where a drone was flying. Low and behold, when the UDS operators reviewed their footage, they found that they did indeed momentarily pick up video of the intruders—only they hadn’t spotted it at the time. “It could be numerous reasons—the operator looks away from the screen for 20 seconds, or goes to grab a cup of coffee and misses it,” said Cedric Coetzee, general manager for rhino security at Ezemvelo KwaZulu-Natal Wildlife. “It’s not going to go beep-beep-beep and off you go. You have to spot it.” The footage is now being used as collaborating evidence in an investigation, but had the pilots spotted the intruders while they were in the reserve, there is a chance that rangers could have apprehended them on the spot.

Otto and his partners believe this problem can be solved with a technical fix. Rob is creating proprietary software that will analyze footage for relevant detection events like a person or a dog in the park, while Serge Wich, founder of the non-profit group Conservation Drones and an ecologist at Liverpool John Moores University in the U.K., is also working on an automated solution. Wich has a graduate student assigned to the project and is getting additional help from researchers in the astrophysics department—the unexpected result of a conversation he struck up with a colleague on a commuter train. “In a way it’s an odd collaboration for a biologist, but astrophysicists look for thermal objects like stars that occur against a dark, cold background, so they actually have a lot of expertise in this,” Wich said. A rudimentary prototype that he and his colleagues developed can already detect people and wildlife, but only in open grassy fields. A more advanced version that could work in the African bush should be ready in two years or so, he said.

That either Rob or Wich will succeed in the near future is not certain. As Lucas Joppa, lead environmental scientist at Microsoft Research, pointed out, image processing is still an imperfect science. For every person Facebook’s auto-tagging photo system correctly identifies, another smiling face stumps those advanced algorithms—and that’s with the backing of a fleet of programmers. “All this stuff is possible in the realm of computer science, but these aren’t just functions in Excel that you click and do your annual taxes with,” Joppa said. “You need software that in real time runs computer vision algorithms that can identify a human versus an elephant versus a monkey that looks a whole lot like a human.”

Cost, too, remains a huge obstacle. The WWF support in Liwonde will likely last a year at most, and even with the help of Air Shepherd and a few other donors, UDS is not making enough to keep operations out of the red. Otto and his business partners have invested more than $1.4 million of their own money, or about half of all costs to date. They’re trying to build up the commercial side of their business—using drones to patrol for nocturnal intruders at industrial mining sites, for example—but in the meantime, conservation deficits come straight out of their bank accounts. “I can’t keep that up, I’m not a rich guy,” Otto said. Asking national parks and reserves to pay for the service themselves likely isn’t an option, either; with annual budgets often in the neighborhood of $500,000 to $1 million, park managers cannot shell out for a drone system that costs about $40,000 per month per field team.

Joppa thinks that the tech industry could play a helping hand in offsetting the costs associated with anti-poaching—not by giving out one-off $5 million grants to conservation groups, but instead by donating labor from their software developers and hardware engineers. Technical expertise, in other words, is more helpful than cash to spend on tech. The benefits, Joppa believes, would go both ways. Conservation presents some gnarly tech challenges, calling as it does for power-efficient, ultra-sturdy devices with long-range transmission capabilities and complex yet easy-to-use capabilities—all for an affordable price tag. “If something works well enough that I can carry it around in my pocket, that doesn’t mean it will work in the Serengeti,” Joppa said. “But if it works in the Serengeti, then I can probably figure out how to get it into a form and price that works in my pocket.”

UDS has already produced such innovations; its bespoke BatHawks prove that payload-carrying drones can operate in rugged places for a relatively reasonable price-tag. And should UDS and others continue to invest in anti-poaching drones, the technology will only get better and cheaper. For now, though, the pilots must hold their own with what resources are available.

Calling it a night, Stephan switched the BatHawk back to manual to navigate its landing. We all stood by, looking up and waiting. From the blank black canvas of the park’s sky emerged a quiet buzz, then the red and green pinpoints of the drone’s lights. Closer and closer it came, until it smoothly touched down, skidding to a stop. A nearby group of hippos expressed their discontent at the disturbance with grunts like Jabba the Hutt’s laughter and then slid into the river—for now cleared of illegal fishers.

Sources Cited  

Anon. (2012) ‘Google Helps WWF Stop Wildlife Crime’, WWF, wwf.to/2jbFUG5.

Anon. (2013) ‘Malawi: Statistics’, UNICEF, uni.cf/2xbyYxR.

Anon. (2014) ‘Motor vehicles per 1000 people: Countries Compared’, NationMaster, bit.ly/2eM74OE.

Anon. (2017a) ‘Forest resources: Malawi’, FAO Forestry Department, bit.ly/2vUiuL4.

Anon. (2017b) ‘GDP per capita (current US$)’, The World Bank, bit.ly/2jbX325.

Anon. (2017c) ‘Research Methodologies: Gartner Hype Cycle’, Gartner, gtnr.it/2gRq82g.

Butler, R. (2006) ‘Forest Data: Malawi’, Mongabay, bit.ly/2xjSSHM.

Chonghaile, C.N. (2015) ‘Malawi floods devastation far worse than first thought’, The Guardian, bit.ly/2wfdLik.

Emmanuel, N. (2013) ‘Democratization in Malawi: Responding to International and Domestic Pressures’, African and Asian Studies 12: 415-434.

McNeil, D.G., Jr. (1997) ‘Kamuzu Banda Dies; ‘Big Man’ Among Anticolonialists’, New York Times, nyti.ms/2wNENBH.

Mitchell, M. (2002) ‘Living Our Faith: The Lenten Pastoral Letter of the Bishops of Malawi and the Shift to Multiparty Democracy, 1992-1993’, Journal for the Scientific Study of Religion 41: 5-18.

Morris, B. (1996) ‘A Short History of Wildlife Conservation in Malawi’, Edinburgh University Center of African Studies Occasional Papers: 64.

Nyarko, E. (2017) ‘Malawi’, World Health Organization, bit.ly/2gPo1IL.

Somerville, K. (2016) Ivory: Power and Poaching in Africa (London: Hurst & Company).

Thornycroft, P. (2016) ‘South African ranger arrested for allegedly killing white rhino’, The Telegraph, bit.ly/2jbGbsB.