This report investigates unmanned vehicles with autonomous functionality and methods for increased fail safety with a focus on fault detection, diagnostics and handling. A vehicle's ability to perform during a mission is limited by its fault prevention, avoidance, tolerance and recovery capabilities. These four abilities can be seen as layers of measures and if one layer fails another should take over. The vehicles' level of fail-safety is increased if more layers of measures are added to combat a specific fault. The list of faults constitute risks whose consequences have to be handled in the design and use of the vehicle. An analysis of those risks can inform method choice in the form of technical solutions. An important conclusion from the report is that much of the research on fail safety has been domain specific, leading to divergent taxonomies as well as different frameworks for risk analysis and mitigation. Together with the report "Fail-safety for unmanned vehicles: Domains", this document provides a knowledgebase and a starting point for further research.