Assessing Fatality in AI, ML, and NN Solutions for Critical-Safety Systems in Transportation, Industrial, and Construction Industries

This study explores the trend to apply artificial intelligent (AI), machine learning (ML), and neural network (NN) based solutions to safety-critical systems in transportation, industrial, and construction sectors of the industry. NN is introduced as subsets of AI and ML. The fatality rates for the three industrial sectors are presented as a reference for the expectation of AI, ML, and NN solutions in safety-critical systems. The reliability of such systems has to significantly improve and probability of fatality in such AI solutions to safety-critical systems should be hundred or thousand times or higher order of magnitude lower. Reliability standards and methods for AI solutions have been and are being developed to meet the stringent requirements.