LIFTAN: An experimental expert system for analysis of manual lifting tasks

Publisher Website
Google Scholar
Abstract
A knowledge-based expert system (LIFTAN), designed to aid the analysis of manual lifting tasks, is described. The system, written in VAX LISP and implemented on a VAX-8600 mainframe computer, consists of three modules: a knowledge base, a general-purpose inference engine, and a human interface with explanation capabilities. The system allows a non-expert in the field of manual lifting to utilize the relevant knowledge and apply it to analyse specific work situations. It can also be used as intelligent computer-assisted instruction for engineering students. The knowledge base of LIFTAN consists of a total of 159 rules, with 91 rules related to risk analysis and 68 rules used to create explanations and messages to the user. The first group of rules concerns the task and operator variables and relates their specific values to the potential risk of low-back disability due to manual lifting of loads. The values of the variables used in the knowledge base are represented using the linguistic approach, and therefore allow for the natural imprecision of verbal descriptions of real situations. Two real case studies are shown in order to illustrate the practical implementation of the LIFTAN system in an industrial environment.