This monograph provides a tool-set for hybrid estimation that can successfully monitor the behavior of complex artifacts with a large number of possible operational and failure modes such as production plants, automotive or aeronautic systems, and autonomous robots. For this purpose, ideas from the fields of System Theory and Artificial Intelligence are taken and hybrid estimation is reformulated as a search problem. This allows to focus the estimation onto highly probably operational modes, without missing symptoms that might be hidden among the noise in the system. Additionally a novel approach to continue hybrid estimation in the presence of unknown behavioral modes and to automate system analysis and synthesis tasks for on-line operation are presented. This leads to a flexible model-based hybrid estimation scheme for complex artifacts that robustly copes with unforeseen situations.