Event sequence reconstruction in automated guided vehicle systems

This paper addresses the problem of reconstructing the event sequences in an automated guided vehicle system (AGVs) that is modeled as a Petri net. We assume that every location of vehicles in the AGVs (i.e., each place in the net) is equipped with a sensor that is able to detect the presence of vehicles. Furthermore, the observation of each sensor is asynchronous and each sensor only knows the ordering of its local observations due to the lack of global time. Our goal is to reconstruct the movement trajectories (transition firing sequences) of vehicles based on these asynchronous sensor observations. We develop an algorithm that is able to obtain these event sequences that are consistent with both sensor observations and the Petri net structure.