Research

Prototype Solution for Human-Elephant Conflict using  Secondary Data

Co-Authors - Dinushi M. Samarawickrama , Harshana Serasinghe

Department of Information and Communication Technology, Faculty of Technology, 
University of Sri Jayewardenepura, Sri Lanka

Website - elenotify.xyz

Introduction

Sri Lanka and elephants have been intertwined since ancient times. Sri Lankan elephants are world-famous as a fundamental symbol of culture. The Sri Lankan elephant community, with its proud body, makes a significant contribution to the tourism industry. But the decades-long human-elephant conflict has escalated to the point of loss of life. This conflict has now escalated into the killing of human elephants. The main reason for the human-elephant conflict in Sri Lanka is the loss of land where elephants and humans live, which can be attributed to each other's encroachment on each other's territory. As a result, the elephant distribution in Sri Lanka is 59.9%, of which 69.4% is inhabited. 

As a result of this conflict, the elephant population in Sri Lanka has declined by 16.1% since 1960, and its current distribution is largely declining, with 318 elephants dying and 112 losing human lives by 2020. Not only the loss of life but also the damage here affects the development of the country.

Electric elephant fences have been used since 1966 as a major solution to reduce human-elephant conflict. In 2018, US $ 1,068,021 was spent on all solutions, including elephant fences, radio collars and GPS collars. But these methods have not yet become successful solutions. As a result, Sri Lanka is moving to identify and tracking elephant movements and behaviours instead of harming or repelling the elephants currently in use globally.

Analysis of the literature made it clear that these motion identities occur under three modes of vision-based, sound-based, and seismic-based recognition. The seismic research was presented because of the positive results of many studies on the use of ground motion, the satisfactory contribution percentage of the data available for such research, and the country's ability to adapt to the cost.

Method

The frequency of seismic waves generated by elephant footprints is 4.5 - 80 Hz, which is captured by a geophone. A secondary dataset was used in this study, which was linked to logistics regression, a machine learning model.

The result was a mobile app designed to show the movements and behaviours of elephants. The REST API handles the data in the mobile application. The “accuracy_score” method and the “confusion_matrix” method obtain 93% Accuracy prediction.

Conclusion

Only a handful of technical research, such as seismic waves, still takes place. Researches are hampered by economic problems, data collection, activation of actual environmental research, and lack of necessary equipment.

But this kind of practice, which is universally accepted, may cost money in the beginning, but it will be a viable solution in the long run.

LET'S PROTECT 

THEM AND US