The pervasiveness of cell phones, mobile applications and social media generates vast amounts of digital traces that can reveal a wide range of human behavior. From mobility patterns to social networks, these signals expose insights about human behaviors and social interactions. In this talk, I will discuss approaches that can help local governments and non-profit organizations understand better the spatial dynamics of cities and communities, offering additional insights beyond more traditional sources of information. I will first give a high-level overview of the research that I lead in the Urban Computing Lab; followed by an in-depth description of two projects. The first project will present a novel approach to create cycling safety maps at large scale. These maps are used by departments of transportation to understand barriers to cycling adoption. In the second project I will describe a new method to enhance the fairness of mobility-based crime prediction models by addressing data bias. This work highlights that controlling for under-reporting can improve both the fairness and the accuracy of mobility-based crime prediction models. My ultimate goal is to illuminate the determinants of human dynamics and to understand the role that the context – represented as physical infrastructure and social fabric – plays in people’s mobility experiences, which can in turn assist in the design of more efficient and inclusive cities.