In this chapter, we study techniques and methodologies for cleaning mobility data, which requires specific tasks not present in standard data science applications. In Sect. 9.1 we classify the kinds of mobility data that usually require cleaning tasks. These include static attributes, voyage-related attributes, temporal attributes, and spatiotemporal trajectories. We continue in Sect. 9.2 with basic cleaning tasks over the AIS dataset. In Sect. 9.3 we address cleaning of static attributes, which describe the moving object as a whole, and continue in Sect 9.4 with the cleaning of voyage-related data, which are attributes that remain static during a voyage. In Sect. 9.5 we delve into the cleaning of temporal data, namely, data that change during a trip. The remaining of the chapter addresses the cleaning of spatiotemporal trajectories of moving objects, as reported by location-tracking devices such as GPS. Here we consider two types of movement. Free-space movement, such as a ship in the sea, poses challenges for cleaning due to the lack of predefined routes. Kalman filter smoothing is a typical method to address this problem. In Sect. 9.6 we explain the theory behind Kalman filter and focus on the practical implementation over the AIS dataset using the Stone Soup tracking library. Finally, in Sec. 9.7 we address network-constrained movement, such as the movement of cars on roads. Here, sensor errors can cause the reported points to be incorrectly placed outside the network. This problem is typically addressed using map matching, which in turn, use the hidden Markov model (HMM) as a tool to align trajectory points with the underlying network. We illustrate the implementation of HMM map-matching using a Valhalla server.