Implementation of spatial representation and autonomy in material handling systems

The paper discusses the use of spatial autonomy for detecting and avoiding physical conflicts by using examples from a warehouse environment. Like many other real world environments that are the subject of simulation, a warehouse has numerous autonomous elements, including workers, AGVs and human-operated equipment. Traditional simulations have not fully considered the effects of this autonomy, which can have significant impact upon system performance and result in accidents, delays, and deviation from modeled behavior. In the past, modelers have treated these deviations as inputs to simulations, hidden in travel time distributions and delays based on real world observations. Okashah (1994) examined four shortcomings of discrete event simulation: (1) arc and node limitations; (2) predefined entity goals, behaviors and interactions; (3) variable-frame of reference; and (4) space as a resource. The paper explains an approach to resolve these issues using spatial representation and autonomy. An autonomous approach to simulation results in deviations from idealized behavior becoming an output of-rather than an input to-simulation. Individual incidents are thus detected and resolved in simulation, and avoided in the real world.