Since a few years, we talk about artificial intelligence for our lifts. Today, we would like to clarify the situation.

 

Let’s try to explain what is the Artificial intelligence.

To be simple, Artificial intelligence is a new type of computer program. The idea was born in the 50’ but the machines were not powerful enough to make it work. We had to wait until 1990 to see "Watson" of IBM (an artificial intelligence program) win against a world class chess champion. In those programs, human do not control everything, the software analyse and choose its own actions.

More recently, we proved that the software could learn by itself, it is called the “Machine Learning”. Examples are usually given to the software and from those examples, it can calculate and generate a result.

For less than 10 years, it went even further with the “Deep Learning”. The software can now find the result without any help. For example, there is a software that can determine if a person is young or old. After a lot of recognition, the software was able to say that our ears size is determinant because they are growing all our life.

 

And now, the Big Data

You probably noticed that when you are doing your shopping online or just when you are watching a product, you will be target by advertising right after about the same product.

It is possible thanks to the massive data processing. Here again, in the 60’, it would have took years to process all the data. Today, it can be done instantly and this technology is used for meteorology, medicine or maintenance.

 

Firsts application of Artificial Intelligence

In order to fluidize traffic in high buildings, they developed algorithms of sophisticated maneuvers. For example, if we are aware that at 9 am, a lot of people will be in the 3rd floor to drink a cup of coffee, the elevators will have to be ready for it.

 

Maintenance with Artificial Intelligence

A lot of manufacturers are making alliances with companies like IBM, Microsoft or GE Digital to implement artificial intelligence in their elevators.

 

How does it works on elevators ?

It is all about retrieving data and we can easily find 3 ways for it.

In modern elevators
The control panel can be used for that. Some manufacturers are talking about 300 different data. For example: speed, number of runs, weight, door openings, landings, etc.

In old elevators: remote monitoring
The phone on the cabin can be transform on a remote monitoring system. A connected device can be plug at specific spots to retrieve the needed data.

In old elevators: The Black Box
Another solution from a manufacturer. A connected box is plugged on the cabin roof and it collects data. It can retrieve acceleration, speed, height, etc.
This manufacturer claims that The Black Box can retrieve 200 different data.

Nota: a french newspaper, “Le Figaro” revealed a startup named Uptime. It was founded by Augustin Cellier, a former OTIS employee and his brother. They install two boxes, one on the cabin roof and the second on machinery. According to them, the average for an elevator is about 3 failures per year. Today, they can promise 0 failure. How can they do that ? Simply because they have the information of a defective part before the break and with this informationare they are able to prevent the failure by replacing the part.

 

How does it works for lift technicians ?

Lift technicians are now equipped with smartphones. They can receive their schedule and they can directly reassure the customer and their managers. Thanks to that, we have access to live monitoring on failures and repairs.

 

How does it works for data processing ?

All the data are processed by correlations. If there are a lot of elevators connected, the software will be able to notice that, when the ride time change, there will be a failure right after. It can process a lot of data like this.

 

Last step, Elevator engineers

This is where the data is exploited. When we know that the ride time change, what do we need to do ?
With the right data and the right conclusion, the lift technician knows instantly what he needs to do.

 

First conclusion

The potential of this market is huge. And the 4 big (OTIS, KONE, SCHINDLER and THYSSENKRUPP) intend to seize it.

KONE
The 29th september, 2017, in London. Hughes DELVAL announced that they already connected 100 000 elevators. They found a way to easily retrieve data from all the brands elevators of their maintenance park. They are using “Watson”, the IBM platform.

OTIS
The REM (Remote Elevator Monitoring) is already retrieving data like stop accuracy, doors operation, etc.
The 23rd august, 2017, in Taiwan, they introduced their technological advances. They have 300 000 connected elevators and thanks to that, they can establish health reports and determine parts to be changed.
In april 2016, they announced a partnership with MICROSOFT in order to connect their elevators to the cloud.

SCHINDLER
The 6th july, 2016, they made a partnership with GE DIGITAL. They use the software Predix to process data and prevent failures.

THYSSENKRUPP
They are using the software MAX, a technology based on Microsoft Azure IoT. It is a mix of cloud technology, Big Data and Machine Learning.

The battle has already begun. And here, it is not the Artificial Intelligence that is at stake but the Human one.