Thermal imaging provides a way of measuring the temperature of the subject in a contactless manner and is capable of taking the temperature of multiple subjects at a time. Due to COVID-19, there is a high demand for massive temperature screening for the crowded public in high traffic areas to detect any individual with a fever immediately. In this project, we propose to design an open-source AI thermometer for massive fever screening combining edge-based object detection solution and multi-modal sensor fusion. Different from many recent products that target commercialization and still not accessible for small business owners due to manufacturing delay and high cost, our open source solution can help developers build and assemble the AI thermometer like Lego blocks with off-the-shelf open computing board and components. We hope our solution can help small business owners to quickly utilize our AI technology at the lowest cost without long waiting.