A Humanoid-Robotics Method for Predicting Construction Work Performance

A major obstacle to the refinement of modem construction engineering practice is the inability of the industry to exactly predict (in some theoretical rigorous manner) the before- the-event performance of construction workers in the execution of particular tasks or to account for the actual method or systems of work employed or to allow for small changes in the task or in its solution technology. This paper addresses the (novel) worker-task performance prediction problem and seeks to solve it through the application of humanoid robot programming and simulation techniques. The specific aim of this paper is to report on progress that has been made in the quantitative modelling of complex construction worker behaviour though humanoid robotics and virtual reality prototyping means and to present a proof of concept model of a working system.