As the world is getting instrumented with numerous sensors, cameras, and robots, there is potential to transform fields as diverse as environmental monitoring, search and rescue, security and surveillance, localization and mapping, and structure inspection. One of the great technical challenges is to control the sensors, cameras, and robots intelligently in order to extract useful information. In this talk, I will present a unified approach for autonomous information acquisition, aimed at improving the accuracy and efficiency of tracking evolving phenomena of interest. I will formulate a decision problem for maximizing relevant information measures and focus on the design of scalable control strategies for multiple sensing systems. First, I will present an approximation algorithm for non- greedy informative planning with linear Gaussian models. The approach reduces the complexity in the length of the planning horizon and in the number of sensors and provides suboptimality guarantees. An application to active multi-robot localization and mapping will be presented. Next, I will remove the linear Gaussian assumptions and will address active object recognition and robot localization using detected objects. The techniques presented in this talk offer an effective and scalable approach for controlled information acquisition.