Exploring the taxonomy of commits in cyber-physical systems for enhanced error fixes investigation

Cyber-physical systems are a symbiosis of multi-level control systems that take into account the physical aspects of the functioning of target objects. Errors in such systems can be associated both with incorrect organization of the code and operation of the hardware, as well as with an incorrect understanding of physical laws and their numerical approximation. Continuing our previous work, we apply technologies for analyzing commits in git repositories of some well-known cyber-physical systems, followed by classification of messages from developers. As a result, we discuss the identified strong keywords and generalized fix messages that can reveal the main classes of bugs in these projects. The results of the work can be used in training and consulting on errors and vulnerabilities in complex systems.