Fauna care

Problem it solves

Our project addresses the critical gap in immediate diagnosis and treatment of pet animals, which often arises due to a lack of awareness or the absence of a proper diagnostic pipeline. This solution provides a comprehensive, accessible, and reliable system for monitoring and diagnosing pet health issues.

How It Helps

Early Detection and Treatment: Enables pet owners to detect and treat health issues early, preventing more serious conditions and reducing veterinary costs. Comprehensive Health Monitoring: Tracks vital signs such as heart rate, blood oxygen levels, and temperature, ensuring continuous health monitoring for pets. Convenient and Accessible: Offers a user-friendly mobile app and website for easy health tracking and diagnostics, making it accessible to pet owners anywhere. Emergency Support: Integrates telehealth services for real-time consultations with veterinarians, providing immediate support in emergencies. Inclusivity: Features an IVR system for users without internet access or smart devices, ensuring that all pet owners can benefit from the service. Data-Driven Insights: Utilizes advanced machine learning to filter noise and analyze symptoms accurately, providing reliable health assessments. Species Classification: Uses image recognition to classify animal species, optimizing symptom analysis and improving diagnostic accuracy. By simplifying and enhancing pet health diagnostics, our solution ensures pets receive timely and effective care, promoting overall animal well-being and contributing to broader biodiversity and heritage conservation efforts.

Challenges we ran into

Noise in Sensor Data: Challenge:One of the major hurdles we encountered was the presence of noise in the sensor data, especially from the electrodes used for ECG monitoring. Solution: We implemented an advanced machine learning model (LSTM) specifically designed to filter out noise, ensuring that the data fed into the main diagnostic model was clean and accurate. Image Classification Accuracy: Challenge: Achieving high accuracy in species classification using image recognition was difficult due to the diversity of animal species and varying image quality. Solution: We employed a neural network (CNN, EfficientNetB0) along with Explainable AI (Grad-CAM) to improve classification accuracy and provide better model interpretability. Integration of Multiple Sensors:

Challenge: Integrating various sensors (AD8232, DHT11, MAX30102) and ensuring seamless communication between them was complex. Solution: We used Arduino IDE and C++ for precise sensor control and data acquisition, and conducted extensive testing to ensure reliable integration and performance. User Accessibility in Remote Areas:

Challenge: Providing access to users in remote areas without stable internet connections or smart devices was a significant concern. Solution: We developed an IVR (Interactive Voice Response) system to deliver guidance and support in local languages, ensuring inclusivity and accessibility for all users. Real-Time Telehealth Integration: Challenge: Integrating real-time telehealth features such as video calls and chat with veterinarians posed technical and logistical challenges. Solution : We utilized robust backend services and APIs to ensure stable and secure telehealth interactions, and conducted user testing to optimize the user experience.