Human Factors and Safety Aspects of Intelligent Transportation Systems

Principal Investigator(s):

Peter Hancock, Former Researcher, Kinesiology

Project summary:

The present proposal represents a request for support for Dr. P.A. Hancock, the Director of the Human Factors Research Laboratory for summer, 1995 and a sabbatical year commencing in fall, 1995. It is proposed to evaluate the linear, complex, and chaotic framework of accidents within a temporal context. Briefly, the understanding of accidents presents a fundamental challenge to the scientific method. Science seeks influence through its ability to codify experience and to use such information to predict future events. This tactic has provided the phenomenal success of science in the physical realm where the tools of mathematics and computing have greatly enhanced predictive capabilities. However, within the last decade there has been a revolution in science in which we have gained substantive insight into the indeterminate but 'patterned' world of non-linearity, or more popularly 'chaos'. Even more interesting developments have occurred within the last two years in research on complexity defined as residence at 'the edge of chaos'. These advances promise the opportunity to integrate the uncertainties and nuances of human behavior into a common quantitative framework and represent the future of human factors as the science of human-machine systems. Nowhere will this integration be tested more thoroughly than in an understanding of accidents. Typical failures of human-machine systems have predominantly been ascribed largely to 'human error' where human error is described, after the event as some vital act, committed, omitted, or incorrectly sequenced. Thus there is no 'science' of human error per se. There are descriptions of error, which one central function of epidemiology. There are imaginative strategies which seek to protect from potential error. However, there is no fundamental approach which addresses the prediction of individual accident events. Is such an approach feasible? What is the 'grain' of analysis in which 'prediction' becomes possible? How

Project details:

  • Project number: 1995030
  • Start date: 06/1995
  • Project status: Completed
  • Research area: Transportation Safety and Traffic Flow
  • Topics: Safety, Traffic operations