Precision Livestock Farming


Winner Project in Locus Event 2019, Under Hardware Thematic Category
2019
Poultry is one of the contributing areas for the growing problem of antimicrobial resistance in Nepal. Broiler chickens are prone to mass casualty and disease outbreaks with no significant symptoms leading to huge economic loss for the farmers. Moreover, they are made to grow fast, often times unable to support their own weight causing leg disorders and paralysis.

This project presents a novel method that monitors the mobility, distribution (using images from camera processed with trained YOLO and SORT algorithm) and feeding behaviour (using sound analysis) to estimate the welfare of the fowls while providing suitable habitat for growth (with environmental sensor and ac appliances). This results in increased profit for the farmers. We also found the broiler to prefer high temperatures and humidity, and good ventilation a must to curb the effects of methane and ammonia gases produced from the excreta of the chickens. Poultry is one of the contributing areas for the growing problem of antimicrobial resistance in Nepal. Our project can also help prevent outbreak of diseases and introduce timely intervention that can minimise unnecessary antibiotic administration.
Tracking hens in order to see their
                                distributution

Tracking hens in order to see their distributution

Precision Livestock Farming (PLF) is data Data Collections & Data Analysis project that uses various advanced technologies to monitor and understand animal behaviour in order to optimize their contribution to the livestock. Our prototype monitors broiler chickens in the poultry using camera, microphone and various environmental sensors to examine the fowl behaviour (feeding, mobility, distribution, etc) which has been correlated to the well being and sound health of the broiler variety.

The user friendly mobile application has been developed to monitor the status of poultry farms and notifies the farmer in critical situations. The app has got the features to control hardware appliances such as heater, light, humidifier to maintain suitable environment for them.

Camera is used to track each chicken for their distribution and mobility index. Dense distribution implies ventilation failures, illumination changes and feeding problems and low mobility index reveals locomotive issues, lameness (paralysis) or leg disorders. Microphone is used to detect peck counts and quantitatively estimate feeding behaviour. Other sensors (temperature, humidity, air quality and light sensor) are used to maintain suitable habitat, assisted with heater, fan, light bulbs and other connected ac appliances.
Arduino Shield &
                                                        Arduino Uno
Prototype Circuits
                                                        for Sensor Data
                                                        Accumulation
                                                        and
                                                        Actuations
Atmega32, Display,
                                                        Buzzer and Indicator
                                                        Board
Project consisted of major dissections - Data fetch using Sensors, Analysis and Control via Mobile Application. My roles were solely in Mobile Application, Server Setup in Raspberry Pi and Database handling with few contributions to other sections.
Server Setup
Raspberry Pi has been used as interface to support bidirectional hardware and mobile app data flow. Sensors such as camera, microphones generate provides the data which is passed to Raspberry Pi and result is published to Arduino with UART, other hardware appliances connected. All these sensor data were able to stream near real time in mobile app through python Flask based server that acted as REST API server.
Database Storage
Along with Flask backend server, database was setup in Raspberry Pi to store partial data required for transmission and to enable streaming of derived metrics.
Mobile App
MIT App Inventor was used for user friendly app that streamed near real time data display of sensors. App featured the alarm system incase of emergency situtation in farm. It consists of visual readings, graphs for farmers to observe along with control panel area to control the hardware appliances to manage the suitable environment for the poultry.
Mobile App Screenshots
Winning
The project was demonstrated in 16th Locus event,the most reputed national level tech events, and chosen as Winner of Hardware Thematic Category

Team Members

nischal
rashik
sajil
shrey