The increasing demand for data and smart solutions is one of the fastest growing sectors of human activity. In recent decades smartphones and mobile phones have become a significant and stable source of data. Architects and urban planners have used them in various cases to identify urban patterns. This paper focuses on data gathered by fitness tracker applications which collect information about the movement of their users. The applications record the trajectory of the movement and detect the mode of transport. They require some basic information about the user (age, weight, height and sex) to calculate their caloric consumption. The data from activity tracking smartphone applications create a data lake that can be transformed into a new data source for the designing of healthier and more liveable cities. Combining the data layers and analysing them further could reveal properties and qualities of life in a given location that would not be apparent without processing these data. By analysing the data, we can observe the current state as well as tendencies in human behaviour over longer periods of time. Through the observation and comparison of physical activity in different urban contexts (topography, size of settlement, density of population, density of infrastructure, quality of public spaces, location, etc.), we can develop new alternatives and better knowledge of the influence the above-mentioned factors have on the life in cities. This paper describes data layer combinations that bring novel insights into the connection of physical activity and urban contexts by using data mining technology based on smartphone applications. The theoretical framework can be subsequently applied to various data sets with certain properties.