Floor selection proposal for automated travel with smart elevator

Abstract

Elevators have been used for centuries to convey material and people, with a history going back to 19th century. Modern elevators as we use them today became widely used some 150 years ago, and regardless of many improvements and technological advancements, the general concept has remained the same. The typical elevator still needs traveller’s input to take the passenger from one floor to another. In this paper we explore the possibility to predict elevator passenger destination floor. For this task we use passenger profiles established through deep learning, and elaborate on the passenger’s trip history to predict the floor the passenger desires to travel. The study is based on a smart elevator system set up in a typical office building. The aim is to provide personalised elevator service in the context of a smart elevator.