Current systems that determine parking lot occupancy to inform drivers of available spaces have issues related to calibration and accuracy. In response, this research is creating a truck parking monitoring and calibration system powered by machine learning. Rather than relying on manual calibration, which is labor intensive, inefficient, and difficult to scale up, this research will monitor and calibrate the installed sensing system by creating a data pipeline that includes deep learning and cooperative AI methods. For the developed system, counting data will be obtained by sensors installed at the entrance and exit of a truck parking lot. A separate video surveillance system will collect ground-truth data. Then real-time parking occupancy will be calculated, and error status will be identified by comparing the results with ground-truth records. The cooperative AI calibration component will generate a calibrated occupancy result, confidence rate, sensing system status, and calibration recommendations. The project will also create a live website to show the status of the installed truck parking sensing system and calibration recommendations.