Confidentiality and Real Errors: A Contradiction?

During industrial software development and deployment, a wealth of data is accumulated, which could be used for the evolvement and refinement of methods and tools for error analysis, statistical evaluation of errors, dynamic handling of errors, and the prediction of faults and failures. Unfortunately, this data is always classified as highly sensitive as it contains customer related information, quality and quality assurance related information and gives insights into internal development processes. There is a need for neutralization techniques to overcome these hurdles