COVID-19 spread regulator robot - An automatic robot that detects face mask, check human body temperature, and hand sanitizes people at entrance point

Aim - From 3 January 2020 to 6:19pm CEST on April 21, 2022, there were 247,336 confirmed cases of COVID-19 in Zimbabwe, with 5,466 deaths, according to WHO. Despite the fact that there are remedies for illnesses and that our civilization has progressed by great strides, the most powerful and effective weapon society has against this virus, which affects not only health but also economy, politics, and social order, is the prevention of its spread. This research aims at developing an automatic robot system that enforce minimum spread of the virus by checking body temperature, hand sanitizing people, and check proper wearing of face masks before giving people access to restricted premises. Subjects and methods – The research made use of ultrasonic sensor for hand object detection, MLX90614 sensor for contactless temperature checking, submissible pump to dispense hand sanitizer, and solenoid lock for automatic locking and unlocking of the entrance. This research has built a python model that detects the presents and proper wearing of face masks using MobileNetV2 algorithm and implement the model using OpenCV. The model was implemented in the bigger hardware system. Temperature sensor was calibrated using normal contactless temperature sensors used in public. Results – Hand object detection was successfully achieved with 100%, temperature checking was successful with a variance of +/-5 degrees Celsius, sanitizer dispenser worked 100% correct, and auto un/locking was also 100% accurately functional while the face mask detection model had 98% for accuracy, F1-score, precision and recall. In actual functionality of the system, the model could correctly detect proper wearing of face masks for 90 people in a total of 100 people. Conclusion – Therefore, the model proved to be fully functional with high accuracy score as compared to others that make use of Convolutional Neural Networks (CNN). The model could also be implemented for full functionality in a robotic system that works with zero human intervention. The system benefit communities and organizations attempting to limit the spread of COVID-19 because the control will be done by machines, putting personnel at a lower risk than personally inspecting people at access points. Journal of Technological Sciences, Vol. 1 No. 2 (2023): Zimbabwe Journal of Technological Sciences (JTS)