A Robotics-Based Surveillance System for Livestock Wellbeing and Early Disease Detection in Poultry Farms

This paper presents a smart surveillance system using robotics for analysing the wellbeing of the livestock and early detection of disease in poultry farms. The current system often identifies infections too late, resulting in mass culling of flocks. The proposed solution is to use disease control through vaccination as an alternate approach. The proof of concept for this paper was implemented using the Raspberry Pi 4 along with Arduino UNOs, which was adapted to better suit the needs of the paper. The system was able to collect data from sensors measuring variables such as Ammonia, Carbon Dioxide, temperature, and humidity. The wellbeing of the chicken can be monitored through the environmental variables being sensed. Based on studies, deviations in these variables from the baseline can indicate the presence of disease. Some systems also listen to sounds to understand if illnesses are prevalent. The data will be passed through a series of conditions to determine the probability of infection within the farm. Experimental analyses through simulations were conducted with the mobile robot connected with sensors. Based on the significant deviation of sensor outputs from the benchmark values, the presence of disease in the livestock was detected in the early stages.