New Evidence on Lift Passenger Demand in High-Rise Office Buildings

The planning and selection of passenger lifts for a prospective building relies on requirements on peak traffic patterns, which are usually expressed by peak passenger demand in conjunction with a traffic mix. These requirements mostly determine the size of the lift installation and should therefore be realistic to ensure proper passenger service without excess capacity during the whole life cycle of the building. Despite their importance, real-world surveys on peak traffic patterns are still scarce in literature. In such surveys, human observers typically record passenger demands, but an automated method is required to obtain data on a larger scale and to contest the current requirements. This paper proposes an algorithm that automatically recognizes peak periods occurring in an office building during a day from people flow data, as well as computes peak passenger demands and traffic mixes for the peaks. The challenge is first to recognize the start and end time of a particular peak period, and second, to scale the observed peak demand to the actual population using the lifts during the day. The scaling is crucial when comparing the measured peak demand to the required peak demand, which is usually expressed as a percentage of population. The procedure is developed using measurements of two lift groups in a high-rise office. It is then applied to measurements from other offices to recognize trends in peak traffic patterns. The observed results are then contrasted with the current requirements for planning and selecting lift configurations.