Data warehousing massive real-time elevator signals and maintenance records

Elevators are regarded as important modern day necessities for urban communities. Although elevators are widely used in various installations, the actual control and monitoring of these systems is given less attention by the research community. Specifically, the problem of maintaining and analyzing massive real-time elevator control signals has been largely ignored by data mining community. Every elevator consists of at least one onboard controller (circuit board). Like any other realtime systems, an onboard elevator controller can generate a large volume of signals. The size of real-time signals and maintenance records generated annually from a number of high-rise housing estate elevators can reach hundreds of gigabytes. In this paper, we describe a data warehousing approach for managing massive real-time elevator signals. Our prototype system shows significant reduction in storage requirement and allows efficient query processing across massive real-time data.