Location-sensing mobile devices such as smartphones allow citizens today to produce spatial information - known as Volunteered Geographic Information (VGI). Individuals regardless of their knowledge, cultural backgrounds, or interests contribute, maintain and share these geographic information that comes in various forms such as text, images, or maps. Due to its large potential, VGI data has been used as a data source supporting a wide range of services during the last decade: environmental monitoring, movement analysis, disasters management etc. However, VGI also comes with varying levels of uncertainty due to heterogeneous contributors with heterogeneous purposes using various technologies with different levels of details and precision. Thus, VGI uncertainty measures consist of a combination of context-, usage-, and format-based measures. Furthermore, the lack of sophisticated methods for data fusion hinders the full potential of VGI utilization. To explore the potential of VGI within this SPP call, we propose to develop visual analytics methods to semi-automatically extract and fuse different VGI, and to assess as well as aggregate different sources of uncertainties to solve problems in a crowd mobility context. Crowd mobility is a research field with various fields of study (demography, spread of diseases, etc.). Three main forces drive mobility of humans within space and time: attendance (decision to attend an event at a certain time and place), existence (for day-to-day survival, e.g., going to work), and imitation (follow someone else, e.g. for leisure). The goal of this project is to develop methods to uncover human mobility within these three forces using different VGI sources, and use these findings for solving problems that occur due to crowd congestions in cities, e.g., social stress, anxiety disorders. Furthermore, we strive to develop methods to assess the inherent source uncertainties and propagated uncertainties in fused VGI to increase the accuracy of mobility findings. These findings will aid, e.g., urban planners in their actions to alleviate such crowd-induced predicaments. Without proper handling of uncertainties, we can neither derive new trustful insights nor present the found patterns in a meaningful way. As fully automated systems cannot fully estimate uncertainties, the user has to be involved in the data analysis process. Visual Analytics is a technique combining the strengths of computers with human skills (creative thinking and large background knowledge). Following the Visual Analytics pipeline will enable us to deal with data uncertainties, include domain knowledge and derive new knowledge from heterogeneous VGI data.