On-Line Assistance for Crane Operators

Abstract

Safety is considered to be the most critical issue in the success of any lift operation. However, crane accidents involving tragic deaths of innocent people and destruction of property do occur. The number of accidents has decreased, only after the implementation of strict regulations by the Occupational Safety and Health Administration (OSHA). Repeated boom failure accidents of truck cranes used by the Bridge Maintenance Unit from the North Carolina Department of Transportation (NC DOT) triggered the research study described in this paper. Throughout the summer of 1994, the Bridge Maintenance Division of NC DOT recorded four accidents involving serious damage to their fleet of 64 National Series 446A truck mounted hydraulic cranes. The accidents involved bending of telescopic booms as well as turret drive gear failures. The repair of a bent boom alone can cost up to $16,000 in replacement parts, without considering all the costs of labor and overhead. The main objective of this research project was to develop and test an intelligent monitoring system that could be used to retrofit existing crane hardware. It was shown how a set of sensors and special devices were built and electronically integrated using an analog/digital converter and a laptop computer. Experimental work with a fully operational crane provided the necessary data to build a feature-based state identification system. Using a set of data collected during the experiments the paper demonstrates how sensory patterns can be recognized that lead to the identification of dangerous states.