Going Up with IBM Cloud Foundry: KONE Aims to Fix Elevator Issues Before They Arise

February 23, 2018

KONE is a global manufacturer of elevators and escalators, with a history of more than 100 years. Its equipment can be found in more than 400,000 buildings—including major facilities such as the Staples Center in LA, which hosts 250 events and four million people a year.

One difficulty in the elevator and escalator industry is that traditional equipment maintenance is mostly calendar-based. This means that the components are serviced and replaced according to estimated times / dates of breakdowns. However, these schedules are based on assumption rather than reality, leading to malfunctions that may occur between regularly scheduled check-ups. Aiming at ensuring the safest continuous and efficient “people flow,” KONE looked for a means to address these technical issues in advance of a mechanical failure.

With immense amounts of data collected from its globally-installed equipment, the company realized it could recognize malfunction patterns using AI and take early action. For this purpose, KONE utilized IBM Cloud Foundry and the Watson IoT analytics platform as part of the offering.

With Cloud Foundry underneath, the system now tracks 1.1 million “connected” elevators and escalators around the world. For instance, an elevator is monitored across 200+ performance parameters—such as lighting, time it takes for doors to open and close, vibration, noise level, abnormal stops, humidity, air pressure, temperature, speed, mileage, drive time, load miscalibration and more.

“The key functionality of the system running on IBM Cloud Foundry is predictive maintenance, enabling KONE to supervise its equipment remotely and then dispatch field technicians to troubleshoot malfunctions before it’s too late. “We should never have unplanned call-outs anymore,” says Jaakko Kaivonen, Managing Director at KONE.

Visit the Altoros blog for the comprehensive user story on KONE.

Alexey Khizhniak Profile Image

Alexey Khizhniak, AUTHOR