Researchers have investigated the use of neural networks and reinforcement learning to optimize elevator controls for maximal efficiency, with success in simulation studies.