The design of an efficient elevator operating system

A method is developed for finding an efficient operating policy for an office building elevator system. The method was applied to a particular eleven story building in which there were four elevator shafts. A queuing model was formulated in which the characteristics of passenger arrivals and destinations were time variable. The objective involved the minimization of the weighted sum of ratios of actual to minimum possible travel time between all pairs of floors. Simulation was used to analyze several logical routing policies for each of two methods in which demand information was used to alter the elevator route. The best policy was found to be almost twice as efficient as most of the other policies which were studied and over 25 times more efficient than another seemingly logical operating policy.