Wildlife organizations are trying to use new technologies to track vast areas with a small number of people. Particularly focused and desperate these efforts are becoming in Africa. Here, the illegal trade in animals threatens to destroy rhino, elephant, lizard and lion populations. IoT helps get data to intercept poachers before they reach their victims.
Using sensors to track animals is no longer something unique. For example, Vodafone thus traces seals in Scotland and endangered dugongs in the Philippines. But in Africa, animal welfare has a slightly different specificity: for this, the situation of people is monitored here.
According to the Great Elephant CensusThe elephant population in Africa declined by 30% between 2007 and 2014. These are 144,000 animals. Now the number of elephants is reduced by 8% per year. The main reason - the hunt for poachers for ivory.
The position of the rhinos is even more dismal. The black rhino is no longer endangered (in 1995 there were 2500 individuals, today there are 5000). But the western black rhino was declared extinct in 2011. The last male of the northern white rhino died in March of this year. In total, more than three rhinos per day are killed in South Africa. If the trend continues, they will die out in this area by 2025.
Eyes, Ears and Clouds
For several years now, the Air Shepherd project from the Lindbergh Foundation has been using drones to protect rhinos in South Africa. With the help of crowdfunding program expanded to protect elephants in Malawi and Zimbabwe.
Images obtained from drones are processed to track changes on the terrain model. Animal defenders analyze the movement of animals, the proximity of poachers to them, and warn the Rangers in advance about the threat. But the project requires the participation of trained pilots and a large number of drones.
Therefore, in the South African Reserve Welgevonden Game Reserve tried a different approach. Here, using sensors, they track the behavior of herd animals: zebras, impalas, and gazelles. The project is implemented by the University of Wageningen (Netherlands) and IBM. The movement of animals is tracked with the help of collars. Data is transmitted over the 3G network to the IBM Watson cloud system. IoT platform analysts are trained to identify differences in the behavior of herds of animals during their interaction with natural predators, tourists or poachers.
The IBM Watson solution has so far been effective, but it depends heavily on connecting to the cloud. And for most regions of Africa, an internet connection is not guaranteed.
Consequences of meeting poachers
We go on the other side
An alternative approach is offered by Connected Conservation . The program was launched in 2015 with Dimension Data and Cisco. On the territory of a private reserve for rhinos in South Africa, the position of not animals, but people is monitored. The approach was successful, and it will be expanded to other conservation areas of the continent.
The Connected Conservation program uses fixed cameras, sensors and a radio network to track every person in the reserve. Based on the detected activity patterns, alerts are generated and rangers are sent to intercept poachers. The cloud component is present, but its role is limited: the local server connects to Microsoft Azure to back up critical data.
“When the project started, there were very few communications in the region,” says Wolf Steennes, Dimension Data solution architect. “All communications went through a person on the spot who monitored a laptop and a radio.”
One of the first components installed in December 2015 was the reserve area network (RAN) network based on radio communications. The network with a capacity of 50 Mbit / s uses stationary towers around the perimeter of the reserve, transmitting data from cameras, warnings from sensors and voice messages. The network topology takes into account a number of potential problems, the most serious of which are torrential rains, thunderstorms and constant heat. Ordinary cellular systems could not be used due to the unreliability of GSM in such conditions.
In addition to the backbone radio network, wired local networks were installed at all four automobile gates of the reserve. Video cameras, biometric scanners and remote network monitoring systems, routers and server infrastructure for managed services were installed at the gate. The local IT infrastructure allowed the Connected Conservation team to quickly deploy test equipment and software, which they will also implement in the following program sites. Finally, a Wi-Fi network was deployed throughout the reserve to provide mobile access to sensor data. Through the use of LoRaWAN, sensors can run on batteries for many years and exchange data in the 15 km range.
Two phases of system implementation
Other system additions include thermal cameras installed along the perimeter and fiber optic acoustic sensors. Each vehicle that enters the reserve is equipped with a tracking sensor. Tracked every person who enters the reserve through the gate. Biometric systems scan fingerprints of personnel, rangers, contractors, and suppliers. Visitors scan passports. Cameras record the numbers of each vehicle, and this information is checked in the national database via a VPN connection — the system obtains information about vehicle owners and history of visits. All gates are connected to the control center via 24 video screens. Manage systems trained reserve staff.
Data from cameras and sensors are transmitted to analytical systems. Those generate alerts about abnormal activity or possible perimeter disturbance. Drones are sent to quickly retrieve images, and an armed Ranger team is delivered by helicopter to intercept poachers.
The system proved to be very effective. Since 2015, the number of rhinos removed from the reserve has decreased by 96%, and in 2017 there were no such cases at all. The warning system reduced the average response time of the Rangers from 30 minutes to 7. The number of invasions into the reserve decreased by 68%.
Since the pilot project was successful, Connected Conservation will be implemented in parks and reserves in Mozambique, Zambia and Kenya. Each of these places has its own characteristics, so the technology needs to be adapted to the specific environment and threats. Elephants mostly live in Zambia and Mozambique — and herds are destroyed here faster than anywhere else in Africa.
The next place to introduce the system will be a park in Zambia. The size of the elephant population is relatively stable. Photographing in 2016 recorded 4.2% of dead elephants in relation to the entire population (21,758 animals). But things are very bad in the Sioma Ngwezi National Park on the south-western border of Zambia. There, the proportion of dead animals was 85%.
The park where the system will be introduced is different from the first. There is a big lake where the locals go fishing. But the pond is also used by poachers to track down elephants. Considering the landscape, the physical perimeter fence is impossible.
Therefore, instead of a real border there will be a virtual line of fencing. It will be created by a network of stationary heat chambers on radio masts. There will also be cameras at the entrances and exits. The analytical system will create warnings about the movement of boats through the virtual fence and any night movements in the park.
The authors of the project are working with the Zambian authorities and the fishing community to create a centralized digital access system for fishermen. It will allow to identify those who under the guise of fishermen is in the park for hunting.
The authors of the project dream of its maximum distribution throughout the world - wherever poachers operate. But this requires the support of governments and private organizations.