Elevator IoT Predictive Maintenance Data

This dataset provides time series data for elevator predictive maintenance, originating from the Huawei German Research Center. Its primary purpose is to support the prediction of elevator door failures, thereby assisting in reducing unplanned stoppages and extending the operational lifespan of equipment. The data was collected from various IoT sensors, including electromechanical sensors (door ball bearing), ambiance sensors (humidity), and physics sensors (vibration), during high-peak and evening usage hours in a building. The dataset aims to facilitate predictive maintenance in the elevator industry.

The dataset is provided in a CSV format, specifically as 'predictive-maintenance-dataset.csv', with a file size of 8.54 MB. It consists of 9 columns. The data contains approximately 112,000 records, with some columns, like 'vibration', having slightly fewer valid entries due to missing values.

This dataset is ideal for developing and testing models for elevator failure prediction and predictive maintenance. It can be used to identify patterns indicating potential equipment malfunctions, optimise maintenance schedules to prevent unplanned downtime, and maximise the life cycle of elevator components.

The data is comprised of time series observations sampled at 4Hz. It captures operational data during specific periods of elevator usage, namely between 16:30 and 23:30, in a building environment. There are no specific geographical or demographic scopes mentioned beyond its origin from a German research centre.

CC0: Public Domain

This dataset is valuable for data scientists, machine learning engineers, and researchers focusing on industrial IoT, smart building technologies, and predictive analytics. It can also assist elevator manufacturers, maintenance service providers, and building management companies in implementing data-driven maintenance strategies.

Original Data Source: Elevator IoT Predictive Maintenance Data