There have been more and more accidents in the world. Since 1919, numerous accident causation theories (models) have been proposed. However, due to the complexity of previous accident causation models, it cannot avoid accidents effectively. The liners cannot understand the logical relationship between these causations and apply them in practice. The author and the group think that one of the methods used in preventing accidents is to enable people to remember the cause of an accident (namely, human errors) and do the right thing. We offer 24Model (human‐thinking, simple and easy‐applying) to conduct people to do the right thing for avoiding the accidents. While using the 24Model to explicitly analyze the causes of accidents and then store them in a database in split fields, the statistical law governing accident causation can be deduced. To prevent accidents, Gui Fu and the group call for open data of accident causations for the future preventions. The accumulation and sharing of safety knowledge could promote the safety behaviors, and thus create a safety culture that safety is about doing the right thing. Keypoints: (a) The three kinds and source of unsafe acts directly triggering accidents and disasters are described. (b) The model is a simple, but comprehensive, analytical framework for integrating the findings from accidents. (c) The model can be used to analysis the causes of accidents, form a logical body of accident causes and store them in a database in split fields to enable people to remember and know them. Precis: The unsafe acts arise from the lack of cognition and assessment on hazardous situations, such situations can be improved by reinforcing the safety knowledge of the employees to identify hazards. © 2019 American Institute of Chemical Engineers Process Saf Prog: e12044 2019