This analysis aims to analyse Parking Sensor Data in the city of Melbourne for the year 2019. Given that Parking sensor data is IoT sensor data and consists of measurements recorded across space and time, we use spatio-temporal statistics to predict parking durations and find anomalous events (outliers) within our predictions. An anomalous event is simply a large scale deviation between expected parking behaviour (predicted) and actual parking behaviour. The Kriging model utilised to predict parking durations had a low MSE, however, the model requires heaps more independent variables to enhance our prediction results. Nonetheless, the work explores the application of Spatio-temporal statistics in making predictions and anomlay-detection.