During the average day, three rhinos in Africa are killed by poachers. Most use sophisticated modern weaponry to capture their prize: a rhinoceros horn, which is exported to willing buyers around the world for its perceived medicinal value. Though the horn is only made of keratin, a substance similar to a human fingernail, this hasn’t deterred poachers. In the past decade, 7,000 rhinos have been slaughtered, and last year alone, over 1,000 were killed in South Africa. “The number of poachers is growing rapidly every single day and what that means is that there are more poachers out there with more advanced weapons,” says Bradley Schroder, CEO at the Welgevonden Game Reserve in South Africa. “We’ve had to do the same and this is not sustainable.”
But soon, zebra, wildebeest, eland and impala may actually help save rhinos with the help of an innovative new solution: a fitted, camouflaged, leather collar, which will transmit details about their movements to park rangers, who can then respond quickly. “We detect specific behavior in zebra and wildebeest when they sense threats, be it predators, vehicles or people on foot with rifles,” says Petrus Hendrik Greyvenstein, IBM Executive IT Architect. “If this pattern is recognized in the data, an event will be sent to the wildlife reserve’s command and control center and a response unit can be quickly dispatched. The system provides for a proactive rather than the traditional reactive response. The animals of prey therefore serve as an early warning system.”
This proactive approach stems from the research of Wageningen University. One of top educational institutions in the Netherlands, the university features an animal sciences group focused on research and education related to the health and welfare of animals and people. The solution has been deployed over the last year and is currently being tested in the 38,200-hectare Welgevonden Game Reserve in South Africa. More than 100 animals have been fitted with the collars, which transmit information such as the animal’s location, ambient temperature and their movements. Various experiments are currently underway, whereby data is collected about the animal’s behavior towards different threats in its habitat. A number of specific patterns have been discovered, including those of migration and feeding habits.
The program was initiated by MTN, a multinational telecommunications group. In creating what MTN calls the “Connected Wildlife Solution”, they sought out IBM’s Smarter Networks and Internet of Things (IoT) technology platform. The result is the Connected Wildlife system that captures data, which can be sent both locally to the park and to Wageningen University in the Netherlands. “The first challenge for us was to deploy a Low Powered, long Range (LoRa) Network, using the existing MTN 3G/4G network to back-haul the data collected. The aim was to provide a network to collect the animal data in areas where there was little to no 3G/4G coverage,” says Greyvenstein. “We also had to consider battery life for the sensors on the collars as it is very costly to retrieve and replace these."
LoRa works in dense, hilly areas that are common to game reserves and national parks. In so-called “dark spots” where Bluetooth, WiFi and 3G cannot transmit a signal, cell towers equipped with a solar panel and a battery pack are placed at strategic points where good 3G/4G coverage exist, known as the LoRa Gateway. Animal movements can then be detected via their LoRa Sensor embedded in a collar, and transmitted over the LoRa Network to the MTN Cell Tower using a SIM card, which is then transmitted to the MTN Data Center.
That data is then sent to the IBM IoT Platform. The data is localized on the IBM Platform where it is made available to a number of different solution providers. This information is then used to display animal movement on a dashboard at the Welgevonden Control Center. The second set of data is collected by Wageningen, where an established animal sciences group studies the meaning behind the animals’ specific reactions.
In addition to protecting rhinos, the technology may also be used in the future to help farmers track sheep and cattle, analyze soil quality and help in the mining industry. “This project is providing real business solutions to Africa,” says Greyvenstein. “What we’ve done is developed the analytics, the sensors, the networks —every aspect— it’s a real end to end solution where we’re pushing the boundaries at all levels.”
IBM envisions several phases for this project. The first phase is the collection of enough data to clearly determine how the collared animals react when confronted with any sort of predator, be it people on foot, vehicles, or animal predators.
There are plans to expand the concept throughout Welgevonden deploying to an additional 2,000 animals or more, before expanding to other game reserves in South Africa. Ultimately, the aim is that parks throughout Africa can easily pinpoint detected animal patterns and be able to react faster to possible threats before a rhino poaching incident takes place. “The greatest achievement would be that we can start reducing the number of rhinos being killed, and actually be able to bring poachers to justice,” says Greyvenstein. “We would see rhino numbers increasing instead of decreasing.”