Major failures of technical systems show how strongly the current world and the preservation of our environment depend on the smooth operation of the systems. Events obey that system-design did not anticipate effects of unforeseen conditions leading to lack of safety-functions of a system. However, there are methods obligatory to show that systems can perform safely and to ensure safety functioning in real life conditions, which need to be conducted for any technical system, i.e. predictive risk assessments. As accidents continue to happen, obviously accidents show also the incompleteness of such predictive risk assessments. This paper will provide a systematic for grouping the entire scope of human issues ranging from classical human reliability to errors of commission and will span up to resilience. From this general systematic, requirements for the safety culture and safety in the planning phase of a system development will be drawn, using a safety scanning approach. From this systematic, a logic of human reliability, assessments for current and future systems will be outlined, including how to assess the resilience of systems with respect to human reliability. The importance of event evaluations for modern human reliability management will be shown using examples for occupational health and safety issues.
Finally, the paper will describe how predictive risk assessments can be improved regarding the human role on risk and accidents. It will describe the pitfalls and needs to better Human Reliability Assessment, including task- and goal-related human behavior in risk assessments. The paper concludes with an outline of the potential and use of big-data algorithms for Human Reliability Assessment for active system safety design and organizational human reliability Management.