Early-warning System to Mitigate Human-Elephant Conflict through Acoustic Monitoring

India has an estimated population of 28,000 elephants comprising >50% of Asia’s elephants. Annually, over 400 people and at least 100 elephants are reported killed in India due to human-elephant conflict (HEC).

Rural communities living around HEC-prone regions are often forced to invest heavily in mitigation strategies, adding considerable financial and psychological burden. Barricades and active measures, such as electrified fencing and lighting are often expensive or ineffective when homemade. Many spend time keeping watch at night over their crops. Response to active incidents is often erratic and uncoordinated, resulting in further injuries to both people and wildlife.

In response to the rising reported incidents of human-elephant conflict in the country, the Wild Seve program was set up to monitor and address all conflict incidents around the protected areas of Nagarahole and Bandipur in Karnataka, India. The Wild Seve program currently has helped report over 16,000 incidents of human-elephant conflict. While crop and property damage incidents are most frequent, occurrences of human injury and death are seen in this landscape. Such incidents erode cultural tolerance towards wildlife and result in retaliation against the elephants.

Early warning systems, with the ability to inform vulnerable people, villages and authorities about active elephant movement, can aid in a coordinated response. It can help alert those nearby from venturing out of their homes and allow concerned agencies to effectively patrol regions. An emerging technique, especially in regions with limited resources is the use of acoustic monitoring. These systems, utilising an array of microphones and sensors, can be used to monitor large regions in real-time. Information from the sensors is processed to isolate and identify distinct communication signals from foraging elephants. This information helps contextualise the location of the elephants and is transmitted to nearby communities, frontline staff, and emergency responders.

Our project will create an acoustic early warning system to detect elephants crossing into human habitation. In an international collaboration between Centre for Wildlife Studies (CWS) and Rainforest Connection (RFCx), a real-time AI-enabled acoustic monitoring system will be established in pilot locations in India for early warning for communities living in high-conflict areas. The network of acoustic sensors will detect elephant intrusion and alert governmental agencies and communities through an app or text message. The project will create a low-cost, field-ready, and easily deployable system that will be field-tested to evaluate its effectiveness. 

The CWS-RFCx collaboration will create an acoustic early warning system to autonomously detect elephant presence in human habitation areas. The project will have designed, tested and openly released an affordable, field-ready, and easily deployable system. It will act as a proof of concept for how such an acoustic monitoring system may be spread to all regions in which HEC exists. 

Our goal is to also engage and educate the public on HEC, and on the technology behind the system. We plan to conduct outreach and publicity campaigns, stakeholder interactions, public workshops and the creation of educational material. Our efforts and results will be documented through articles in the media, as well as reports and peer-reviewed scientific publications. 

Our project will be beneficial to the scientific community through improvement in design of acoustic monitoring systems. It will greatly benefit local communities and other associated stakeholders. It will also be beneficial in improving policy and park management, and further conservation objectives for the region. It could also reduce the strain on the capacity of local government and NGO staff working to respond to HEC cases, such that the community could become more independent in safe mitigation strategies.