Elevators are essential equipment for vertical transportation. However, elevator group control is still not easy because a limited number of elevators are available to achieve efficient transportation under many uncertainties.

Special passengers, such as wheelchair users, occupy more space than general passengers; therefore, they need to wait for elevators longer than general passengers because they must wait until an elevator with sufficient space arrives. Since the moral behavior of ordinary passengers cannot wholly resolve this kind of unfairness, we need a support system to help special passengers.

The widespread use of elevators with cameras and improvements in image recognition technology has made it possible to estimate the number of people in an elevator and the amount of luggage to reduce uncertainties in elevator group control. For example, the system can estimate the number of people waiting in the elevator hall and the number of passengers in the elevator. On the other hand, this information for fairness and efficiency has been open to research.

This study is notable because it aims to make it possible to achieve both fairness and efficiency between general and special passengers by focusing on the amount of passenger occupancy acquired from the camera. In this study, the authors formulated this elevator group management as an optimization problem considering the number of people and the occupancy. The authors solved this problem as a multiagent system to achieve fair and efficient transportation.

This paper has remarkable originality and effectiveness in modeling elevator group control using camera information, considering the fairness and efficiency of waiting times between special passengers and general passengers. Compared to existing methods, simulations under various conditions show that the proposed method can achieve fairness and efficiency for both types of passengers. Moreover, the analysis is convincing, even for situations where camera information is erroneous, such as during congestion. Finally, the contribution is significant to the multiagent field and elevator group control. For the above reasons, this paper is worthy of the Best Paper Award.