Thanks to technology, changing the methods of wildlife conservation

Clive Cookson in London During the dry season in Kenya National Park, a small group of poachers walk along a dry riverbed, aiming to kill a black rhino that can fetch as much as $100,000 on the Asian black market.
The men hide under the tangled trees on the banks of the river, but a hunting alert system set up by the Zoological Society of London keeps their eye on them.
Their guns and knives trigger the hidden sensor of this system designed for instant detection, which in turn activates a camera hidden in a forest.
The image is transmitted by radio waves to a base station, and then via a communications satellite to the park's headquarters, alerting authorities in time to dispatch patrols to catch the gang.
Similar scenarios will soon follow in parks and reserves around the world, as conservation agencies adopt a high-tech approach to their battle to protect the animals.
Some of these agencies weren't able to quickly see the potential of the new monitoring tools, but with the help of companies like Google, they are now adopting them as a way to tackle the issue of poaching.
The World Wide Fund for Nature estimates that the illegal wildlife trade is worth about $20 billion annually, and has contributed to a catastrophic decline in some species.
According to the Living Planet Index tracked by the World Wide Fund for Nature and the Zoological Society of London, between 2019 and 2020, 60 per cent fewer vertebrates (mammals, birds, reptiles, amphibians and fish) now live in the wild than before 50 years, the most severe decline in the tropics.
Although the dramatic decline in the numbers of some well-known species such as tigers, elephants and black rhinos has been halted or even reversed by intense conservation efforts, poachers are still killing while other animals, including wild animals, are rapidly declining. Or cinnamon squamous and many monkeys.
There are many drivers behind biodiversity loss, from overfishing and climate change to urbanization and local pollution.
For biting species, poaching and killing are the biggest factor in the decline, says Andrew Terry, head of the conservation unit at the Zoological Society of London.
“We are particularly focused on addressing the issue of the international wildlife trade, but this is an integral part of our broader conservation efforts," Terry says.
Zoologists have used camera traps to photograph dead animals for decades, but until recently these traps had no wireless connection, so operators had to go to virtually every location to extract the film, and then the electronic memory card, which was often filled with images Useless, showing moving branches or other wildlife that provoked the trap.
“The strange thing is that conservationists didn't decide so quickly to rely on technology," says Eric Dennerstein, director of the Wildlife and Biodiversity Unit at Reserve, a wildlife conservation charity based in Washington, DC. This is an opportunity to make a difference by using a smart camera trap and a calling feature."
Other wildlife conservation bodies, including the Zoological Society of London, have started developing tracking and detection technology at the same time, working with technology companies where wildlife conservation is an opportunity to showcase their expertise.
The real-time tracking systems used by the Zoological Society of London, and the TrailGuard system used by RESELF, are in the final stages of testing and will soon be ready for actual use.
“Overall, wildlife conservation organizations do not have the resources to hire and hire expensive software developers and engineers, so we rely on collaboration with the technology sector,” says Sam Secombe, director of the Instant Tracking project.
Zoological Society of London partners include Google and Iridium, while Resolve works with Intel, Microsoft and Inmarsat.
The AutoML system used by Google, which enables people with limited experience to develop artificial intelligence for specific purposes such as image recognition, is being used in the real-time tracking system, which makes it possible to identify people or animals in real time, through captured images. by the camera trap.
“A successful company depends on its ability to quickly collect, analyze and interpret data, in order to make the best business decisions,” says Secombe. “This also applies to the idea of ​​wildlife advocates using camera trap data. By increasing the speed of image analysis, impact can be achieved faster and more effectiveness".
Remote wildlife parks have little or no mobile network coverage, so the real-time tracking system uses radio transmitters to transmit images to a hidden base base, which are then sent by satellite to the headquarters.
The Zoological Society of London tested the first version of the system by monitoring polar penguins, Canadian bears, Australian night parrots, kenya elephants and rhinos, but it ran into transmission problems, especially in areas with thick foliage.
The team has worked on developing a second, stronger and more efficient version, Real-Time Tracking 2.0, which has undergone successful initial tests in Africa and will undergo more extensive trials in the new year at the Western Forest Complex in Thailand and elsewhere before going live.
The quality of the cameras was also an important issue. As there was no type on the market that met the specifications set by the association, so it developed a 5 mega-pixel camera type for the tracker, with a wide range of focal lengths, that is operated either through an internal infrared sensor that tracks the temperature and movement of the animal, or from Through an external metal tracker to detect fishermen.
“It seems paradoxical that most of the hunting cameras being used by wildlife conservationists are designed for the deer market,” Secombe says.
Although the camera has a powerful computer chip that can run an automated processing system for the captured images, to decide whether it is worth transmitting and sending, this feature will not be used at first, so as not to overburden the system. Instead, the image processing will take place in the cloud after the images are sent.
Another development that will happen in the near future is the integration of acoustic sensors, which are triggered by sounds such as the sound of gunfire, the sound of a saw, an engine or the sound of an animal.
The Zoological Society of London is developing a machine learning algorithm for tracking shots in collaboration with Google.
The TrailGuard system used by Resolve, which incorporates the vision-processing chips provided by Intel in the cameras, is already working on local AI image analysis, where only photos of intruders are sent from people, extending battery life. and reduce transmission costs. The first version of the Trail Guard system, which operated in the Gromiti Reserve in Tanzania last year, tracked down more than 50 intruders, so the rangers were able to make 30 arrests from 20 different gangs of poachers, seizing 1,000 kilograms of illegal bushmeat.
Resolve makes 1,000 upgraded TrailGuard units, 300 in California and 700 in China, to be installed in parks in Africa and elsewhere.
This US organization is proposing to protect 100 parks and wildlife reserves over the next two years, by using the TrailGuard system in the 10 routes most used by gamers everywhere. Once the satellite modems are installed, the number of cameras can be increased to 100 in each park.
The installation cost per park can be as high as $17,000 during the first year, and slightly more in the second year, as future operating expenses for data transmission are around $200 per year, much less than alternative protection measures such as drones to locate poachers or Use of additional guard patrols.
Anthony Dancer, who directs the technology program at the Zoological Society of London, warns that new technology by itself cannot stop the unlawful killings. "Most protected sites all over the world are severely underfunded," he says. "Even if technology is made available, many sites will not have sufficient resources to manage this technology or enough guard patrols to provide a wide range of protection."
Besides hunting for meat, horns, teeth, scales, and other valuable products, people also hunt animals to prevent them from attacking crops or livestock.
Resolve plans to tackle the growing problem of protection by modifying the devices used in Project Trail Guard to distinguish animals from people, for a project called Village Guard.
Camera traps, installed along paths used by large animals that eat or trample crops or attack livestock, will automatically track intruders.
The top five targets are elephants and lions in Africa, leopards and wolves in Nepal, and grizzly bears in the United States.
And then the attached loudspeakers will scare away unwanted animals, by making alarm sounds like the sound of a human scream.
In addition to tracking wildlife threats from poachers to angry villagers, wildlife conservation agencies are leveraging information technology to track elusive animals as they analyze the rapidly growing volume of images captured by camera traps installed around the world.
The Zoological Society of London is using both machine learning and human volunteers to get the job done.
Several projects are underway to identify animals through artificial intelligence.
The largest participatory programme, called Wildlife Insights, is located in Google Cloud and combines the company's machine learning expertise with a range of wildlife conservation groups including the Zoological Society of London. It has been trained to identify 614 different species by taking 8.7 million photos submitted by member organizations, and expects to scale up the project quickly, while wildlife conservationists provide more data. The raw resolution range is between 80 and 98 percent, depending on the image quality and species distinction.
“[Wildlife Insights] is a huge open resource system that will enable people all over the world to manage and analyze biodiversity data,” says Dancer.
At a time when artificial intelligence becomes a more powerful tool than ever, humans will always have a major role in identifying wildlife including hobbyists, as well as experts in zoology.
The 'Instant Track' system used by the Zoological Society is a free citizen science app that anyone with a smartphone can use to identify the animals in the images, and has been downloaded 130,000 times. Collective resource analysis of this species is useful for educating and engaging people, as well as providing direct assistance in identifying animal species.
"You don't need to be an expert," says Ketmozes, project manager at the Zoological Society. "You just have to give your best guess. The result is only when ten people give the same diagnosis to the project leader."
Technology is also helping the people involved in the fight to protect wildlife: the 300,000 to 400,000 patrol rangers who work in parks and reserves around the world, according to the International Confederation of Rangers. A system called Smart (the initials for "spatial monitoring and reporting tool"), developed by the Zoological Society of London and other wildlife conservation agencies, enables rangers to collect and sort data from their mobile devices regarding the locations of animals and humans, Including illegal hackers, in order to deploy fewer employees as efficiently as possible.
The SMART system is already being used in 900 protected areas all over the world. The AI ​​software developed by Harvard computer science professor Milind Tambie will be integrated into the system next year. This system predicts the behavior of hunters, as patrols can be directed to potential places where illegal activity is taking place.
With animal populations under enormous pressure, technology has huge potential to enable conservation groups to use their resources more efficiently in their fight against poaching and the wider illegal wildlife trade.
“We urgently need to innovate, and find new partnerships with industry, governments and academia, to develop new solutions,” says Dancer of the Zoological Society. “In this respect, technology and technology partnerships have the potential to become transformative by enabling us to target our resources more efficiently and effectively. and measure the impact.

Trade tracing..A legitimate approach helps reduce trafficking

The main goal of wildlife conservationists is to stop the killing of animals by poachers, but when they fail, forensic science can help catch criminals involved in the illegal wildlife trade.
Researchers are promoting the idea of ​​using fingerprinting technology to improve the chances of getting visible traces of those who have taken animal parts.
Working with colleagues at the University of Portsmouth, the Society for Zoology scientists are using 'jelly lifters' - small leaves coated with sticky gelatin - to remove fingerprints from cassia scales and other unimportant materials such as snake skins. And then those effects are read through special cameras.
Another collaboration, involving the City of London Police and Kings College in London, has developed a new magnetic powder that enables investigators to recover human fingerprints on elephant tusks, with much better identification than traditional methods. This powder can recover fingerprints after four weeks, saving more time to gather evidence about criminals who handled ivory found by police or customs officials.
Wildlife conservationists are also using new genetic analysis in two ways to investigate wildlife crime. First, if hunters leave tiny traces of their DNA on ivory, horn or any other material, they might be able to be traced back to their genetic fingerprint.
Second, animal DNA extracted from illegally traded materials can be used to trace their geographical origins. This possibility may hold particularly true for smuggled ivory, as scientists create a database showing genetic differences between elephant populations in different parts of Africa.
Scientists at Liverpool's John Moores University have used ultra-sensitive DNA probes to help detect illegal animal materials inside large cargo shipments at borders, ports and customs posts. In a proof-of-concept study published this month, they quickly identified small amounts of nucleic acids from tigers, rhinos and mother cassava. Share it

Add a comment