This is our second post in a series on how Microsoft customers are gaining actionable insights on data by operationalizing ML at scale in the cloud. Based on an IoT (Internet of Things) case study, this post is by Vinod Anantharaman, Head of Business Strategy at Microsoft’s Information Management and Machine Learning (IMML) team.
Urban migration is one of the megatrends of our time. A majority of the world’s population now lives in its cities. By 2050, seven of every ten humans will call a city their home. To make room for billions of urban residents to live, work and play, there is only one direction to go – up.
As one of the world’s leading elevator manufacturers, ThyssenKrupp Elevator maintains over 1.1 million elevators worldwide, including those at some of the world’s most iconic buildings such as the new 102-story One World Trade Center in New York (featuring the fastest elevators in the western hemisphere) and the Bayshore Hotel in Dalian, China.
ThyssenKrupp wanted to gain a competitive edge by focusing on the one thing that matters most to their customers – having elevators run safely and reliability, round the clock. In the words of Andreas Schierenbeck, ThyssenKrupp Elevator CEO, “We wanted to go beyond the industry standard of preventative maintenance, to offer predictive and even preemptive maintenance, so we can guarantee a higher uptime percentage on our elevators.”
Fix it before it breaks – ‘Smart’ elevators
ThyssenKrupp teamed up with Microsoft and CGI to create a connected intelligent system to help raise their elevator uptime. Drawing on the potential of the Internet of Things (IoT), the solution securely connects the thousands of sensors in ThyssenKrupp’s elevators – sensors that monitor cab speed, door functioning, shaft alignment, motor temperature and much more – to the cloud, using Microsoft Azure Intelligent Systems Service (Azure ISS). The system pulls all this data into a single integrated real-time dashboard of key performance indicators Using the rich data visualization capabilities of Power BI for Office 365, ThyssenKrupp knows precisely which elevator cabs need service and when. Microsoft Azure Machine Learning (Azure ML) is used to feed the elevator data into dynamic predictive models which then allow elevators to anticipate what specific repairs they need.
As Dr. Rory Smith, Director of Strategic Development for the Americas at ThyssenKrupp Elevator, sums it up, “When the elevator reports that it has a problem, it sends out an error code and the three or four most probable causes of that error code. In effect, our field technician is being coached by this expert citizen.”
In other words, these ‘Smart’ elevators are actually teaching technicians how to fix them, thanks to Azure ML. With up to 400 error codes possible on a given elevator, such “coaching” is significantly sharpening efficiency in the field.
Hear the ThyssenKrupp story in the customer’s own voice in the video below:
Rather than respond to failure alarms after-the-fact, ThyssenKrupp technicians are now using real-time data to identify needed repairs even before breakdowns happen. The Azure ML predictive models used in this solution are continually updated via seamless integration with Azure ISS, creating an intelligent information loop. These models are expected to continually improve with time as more datasets get fed into the system. Because of two-way flow of data and control, technicians can even put an elevator into diagnostics mode and take actions remotely, reducing the need to travel.
Customers across a swathe of industries are deploying enterprise-grade predictive analytics solutions using Microsoft Azure ML – we make it easy for you to get started today.
By using IoT and predictive analytics in the cloud to increase the efficiency of their maintenance operations and elevator uptime, ThyssenKrupp is giving the world’s burgeoning cities a lift they can rely on.