01/08/2026
With support from the Centre for Species Survival Asian Elephant Rapid Response Grant, Bring The Elephant Home continued implementing the project, Achieving Human-Elephant Co-existence through AI-Assisted, Community-Led Monitoring of Individually Identified Asian Elephants in Kuiburi National Park (KNP), Thailand. During this reporting period, the project made significant progress in expanding AI-assisted monitoring, strengthening community participation, and generating scientific data to support more effective human-elephant conflict mitigation.
A key achievement during the quarter was the continued expansion of the Early Warning System. The monitoring network now comprises 68 camera traps deployed across 34 farms bordering Kuiburi National Park, including 18 camera traps equipped with SIM cards capable of transmitting real-time alerts when elephants are detected. Park rangers from three national park substations continued monthly maintenance of the camera network and data collection, ensuring the system remained operational across priority conflict areas.
The project also advanced the development of AI-assisted individual elephant recognition. During the reporting period, more than 1,800 camera-trap images were prepared for AI model development, including over 1,100 images of tuskless elephants and more than 700 images of tusked elephants.
These images will support collaboration with The TRUST Lab at Duke University to evaluate AI's ability to distinguish individual male elephants based on visible physical characteristics. While the current system reliably detects elephants at the species level using SpeciesNet, the next phase aims to identify individual elephants, providing managers with more detailed information to support targeted conflict mitigation and long-term monitoring.
The operational value of the Early Warning System was demonstrated during the reporting period when an adult male elephant triggered a camera trap alert at 7:43 p.m. Elephant response rangers reached the site within three minutes of receiving the notification and successfully deterred the elephant before crops were damaged. This incident demonstrates how real-time monitoring can improve response times, reduce crop losses, and enhance safety for both people and elephants.
The reporting period also marked an important step towards community-led co-management. On 18 May 2026, the project convened its first participatory workshop, bringing together 17 elephant response rangers, 12 farmers from Ruam Thai and Yan Sue villages, local government representatives, and NGO field staff. Participants jointly assessed the existing elephant response system, identified strengths and challenges, mapped elephant movement routes and conflict hotspots, and prioritised areas for future Early Warning System implementation. The workshop provided an opportunity to integrate local knowledge with scientific and technological approaches while strengthening coordination between communities and protected area managers.
Participant feedback reflected strong support for the project's approach. Based on 31 completed evaluation forms, participants reported high confidence in using technology to receive elephant alerts and report crop damage, with an average score of 4.65 out of 5.
Participants also agreed that the Early Warning System improved their ability to respond safely to elephant encounters, scoring 4.48 out of 5, while confidence in contributing to human-elephant conflict mitigation within their communities averaged 4.39 out of 5. These results suggest that the project is strengthening both local capacity and confidence in technology-supported conservation.
Progress also continued in establishing an individual elephant identification database. Using a 21-digit alphanumeric identification system, researchers have identified 124 unique sub-adult and adult elephants within Kuiburi National Park. Since the start of the grant, field teams have completed 150 survey days and recorded 1,078 detections of individually identified elephants. Each observation includes information on location, group size, group composition, social associations, and movement patterns, creating a comprehensive dataset that can be integrated with camera-trap detections to improve monitoring and management.
Behavioural research also continued throughout the reporting period. Researchers collected 656 focal behavioural observations, representing approximately 9,840 minutes of direct observation. These data are being used to develop behavioural profiles for individually identified elephants, improving understanding of how elephants use protected habitats and surrounding agricultural landscapes. Integrating behavioural information with camera-trap detections will support more targeted, evidence-based strategies for mitigating human-elephant conflict.
The project also continued to strengthen local capacity for community-based wildlife monitoring. Community leaders were introduced to the EarthRanger platform and the Thai-language wildlife detection event form, which will enable safari guides, drivers and park rangers to record elephant sightings and other wildlife observations in real time.
Building on this initial engagement, a larger capacity-building workshop is scheduled for the next reporting period, during which EarthRanger specialists will train approximately 20 local guides and drivers and five park rangers in using the platform to support community-led monitoring and rapid response.
Overall, support from the CSS Asian Elephant Rapid Response Grant has enabled the project to make measurable progress in integrating artificial intelligence, ecological research and community participation into a coordinated approach to human-elephant coexistence. During this reporting period, the project expanded real-time monitoring capacity, advanced AI-assisted individual elephant identification, strengthened collaboration between farmers and park rangers, generated valuable ecological and behavioural data, and laid the foundation for wider community adoption of digital conservation tools. Together, these achievements represent important progress towards developing a practical, evidence-based model for reducing human-elephant conflict and supporting the long-term conservation of Asian elephants in Thailand
