Predictive maintenance is an increasingly popular technology, but do you know why? To find out, let's look at one product: The elevator. It's a perfect example of the opportunities and challenges that arise from harnessing artificial intelligence to implement predictive maintenance.
For some years now, elevators have been able to be connected via a GSM gateway, making them a connected object. Today, thanks to this gateway, it communicates in real time with an Artificial Intelligence interface using 3G and 4G networks, thus entering the world of AI.
This GSM gateway communicates a wide range of data collected by sensors strategically placed on all elevator equipment (cabin, motor, cables, etc.).
This evolutionary technology makes it possible to detect the warning signs of breakdowns and intervene before they occur. This makes the elevator safer and more reliable, and maintenance less costly and more responsible.
A challenge: smart building, a solution: Artificial Intelligence
If we look at the current role of an elevator operator, it’s partly one of maintenance, carried out in collaboration with condominium associations and building managers.
However, we are gradually entering the world of Smart Building, which is destined to redistribute the role of each stakeholder in building management.
In this context, in order to establish themselves as major players in the world of Smart building, leading elevator manufacturers have set themselves the goal of developing maintenance 2.0, giving them the ability to act before breakdowns occur and make elevator use safer.
Traditional maintenance is giving way to predictive maintenance, thanks to the implementation of artificial intelligence. All the data collected on each connected elevator is used to create an AI that analyzes the data to detect warning signs of breakdowns and anomalies before they occur.
At Schindler France, for example, a predictive maintenance center has been set up at the head office in Vélizy Villacoublay, enabling real-time supervision of the entire elevator fleet under predictive maintenance. Signs of breakdown are classified, and the most urgent ones generate an automatic request for intervention to the network of field technicians, who know the origin of the problem beforehand. Less urgent faults, on the other hand, will be dealt with during the next maintenance visit, or directly by operators at the predictive maintenance center, who can intervene remotely on a machine. For example, if the sensor that analyzes engine vibrations detects an anomaly, it automatically triggers the intervention of a technician who will have direct knowledge of the problem.
Artificial Intelligence more than relevant
The implementation of predictive elevator maintenance is made possible by the analytical performance of AI, which is capable of analyzing a huge amount of data almost instantaneously.
What’s more, this technology is capable of relearning an elevator’s operation on a daily basis, based on the diagnostics that certain factors may produce on the health of the device. In this way, the system sharpens its expertise as the amount of data collected increases with time and the number of connected devices.
To bring this project to fruition, KONE has chosen to partner with the IBM Watson platform, a leader in the Internet of Things and data analysis. IBM Watson’s performance has already been recognized in other sectors, such as avionics.
To find out more...
The warning signs of breakdowns are detected with the help of analytical tools that use all the data collected by the sensors. These tools include Machine Learning, the Artificial Intelligence technology that enables data analysis through the creation of automatic learning algorithms.
More generally, Machine Learning is used in all predictive maintenance projects. Ultimately, this intelligent, personalized maintenance will enable companies to make their machines more reliable, increase their availability and significantly reduce maintenance costs and carbon footprints.
Data science is destined to become an integral part of all areas of expertise, which is why it is vital to develop your data science skills in order to understand the challenges and issues of the future.
