A Complexity Assessment Methodology for Programmable Electronic Mining Systems

摘要

Mining, traditionally a low-tech industry, is now utilizing surprisingly complex programmable electronic (PE) systems. The functional safety of PE-based mining systems is an international issue and concern. From 1995 to 2001, there were 11 PE-related mining incidents reported in the U.S. and 71 PE-related mining incidents reported in Australia. These incidents are due, in part, to unprecedented levels of system complexity. The National Institute for Occupational Safety and Health (NIOSH) is addressing this issue of system complexity by conducting research to develop a quantitative complexity assessment methodology based on Normal Accident Theory (NAT). The methodology models the behavioral interactive complexity at the level of system requirements. A graph-theoretical approach is used for creating quantitative metrics from Software Cost Reduction (SCR) dependency graphs. This complexity assessment methodology will help realize simpler, safer systems that will be easier to validate and verify. The methodology will benefit mining and other industries as well.