Against the backdrop of the almost total failure of conventional physical fences to prevent deaths due to human-elephant conflict and elephant-train collision, management of the problem requires novel approaches. The elephant is a giant and intelligent animal, who, as has been observed through experience, cannot be effectively confined to an area by physical barriers, most of which are built without even considering long-established elephant passes. Instead, human habitats and facilities, including highways, should be pre-planned following an overall strategy to suitably assign natural resources to each party such that natural habitats are disturbed as little as possible. Hence, we envisage the need for a virtual elephant fence as part of the aforesaid strategy to warn elephants that a certain area is not appropriate for them to be present in or best for the safety of the herd. The present study involved in-depth experimentation to investigate the effects of multiple stimuli on real elephants based on the known elephant responses to such stimuli. The proposed system is an autonomous system that detects the presence of elephants, generates alarms for concerned parties, and coordinatively and tactically repels the animals back to safer areas. Experiment results indicated that the sound of a drone and the sound of tiger growls can both exert a strong deterrent effect on elephants. The sound of bees and the low-frequency burst yielded similar results in terms of encounter and withdrawal reactions. The resulting insights may lay a foundation for a novel approach toward the management of human-elephant conflict and elephant-train collisions.
Keywords: Human-elephant conflict; Multiple stimuli; Virtual elephant fence.