Welgevonden prey species anti-poaching solution


Statistics for this year indicate that by the end of August 2018 508 rhino had lost their lives to poaching in South Africa. In addition, 58 elephants were slaughtered for their tusks in the Kruger National Park during the same period. These figures released by the Department of Environmental Affairs highlight the harsh reality facing our natural resources and the rangers in their line of duty.

While Welgevonden Game Reserve, situated in the Waterberg region of the Limpopo Province, has an adept security system in place, the reserve has been proactive in finding innovative ways to prevent poaching. Over the long term, the current military-style approach to combat poaching is not sustainable.

“This industry cannot sustain the cost, in lives and in money, of this war,” says François Spruyt, Welgevonden chairperson. “We needed to find a way to make the probability of being caught so high that poachers would stop poaching entirely.”

This attitude sparked a collaboration between Welgevonden, MTN, IBM and Wageningen University in the Netherlands. The team has since developed a radical, proactive solution to prevent poaching – a completely different approach and unlike anything already available on the market.

Instead of monitoring the animal’s vitals or implementing technology to detect gunfire, the solution focuses on the movement of common prey species within the reserve. GPS collars were deployed on impala, blue wildebeest, eland and zebra, and scientists have since made use of the data generated to detect when animals deviate from ‘normal’ behaviour.

impala being fitted with GPS collar
Image by Jessica Oosthuyse

“These animals need to be aware of their surroundings every second, in case of there being a threat in the area. We use these animals as a sensor to understand what is going on in the veld,” says Herbert Prins, a professor in resource ecology at Wageningen University.

Previous research shows prey animals behave differently when at risk of predation. By adopting this concept and refining its parameters, the algorithm is designed to correlate erratic animal behaviour to the location of human intrusion within the game reserve – prey animals perceive humans as a threat. As poachers are likely to come across these animals while trying to poach rhino, the system will warn anti-poaching units of an intruder long before they have reached their intended target.

Over seven months, researchers conducted various experiments within Welgevonden’s predator-free 1 200ha breeding camp. These experiments aimed to mimic the types of human disturbances that can be expected within game reserves, such as game drives and bush walks. While human movement data was captured in the field, GPS data from the animal collars was captured via the low power, wide area (LoRa) network, and sent directly through to the Netherlands for analysis.

scientists patrolling welgevonden game reserve
Image by Jessica Oosthuyse

“We monitored what they were doing and students looked at what distance the animals reacted. Every type of behaviour was analysed,” says Prins.

The final results were astonishing. “We are changing the odds,” says Mariana Kruger, general manager for ICT and Converted Solutions at MTN. “Today, the chances of a poacher being detected in South Africa is 3%. This solution will increase that to 86.4% within a 125m radius.”

“By using technology, we are able to do far more, and outcompete our competitors to ensure the survival of animals,” says Bradley Schroder, previous CEO of Welgevonden Game Reserve. Proactive in its nature, this solution will not only save animal lives, but those of their protectors too. Plans to take the solution further are currently underway. The aim is to test the product on a larger scale, incorporating the presence of animal predators, and increasing the complexity and accuracy of the algorithm.

Written by René de Klerk

Copyright 2018 Safari News. All rights reserved. This material may not be  published, broadcast, rewritten or redistributed. 


Leave A Reply