Otis Elevator CIO: Modern apps and IoT for digital transformation
The chief information officer explains how technology is transforming its core business and related services. This inside look offers important lessons to anyone interested in digital. He also reveals secret truths behind the "close door" button.
Otis designed the first safe elevator when he needed to lift heavy building materials, while converting a sawmill into a factory in Yonkers, New York. He made toothed wooden guide rails to fit into opposite sides of the elevator shaft, and fitted a spring to the top of the elevator, running the hoisting cables through it. The cables still guided the elevator up and down, but if they broke, the release of tension would throw the spring mechanism outward into the notches, preventing the cabin from falling.
I spoke with the chief information officer of the Otis Elevator Company to learn about the company's digital transformation. Marcus Galafassi, the company's CIO, was my guest on episode 292 of the CXOTalk series of conversations with the world's top innovators.
This episode offers an inside look into the hidden world of elevators and the people who build and service them.
The first challenge lies in driving change through its globally distributed network of 33,000 field service technicians. As you read below, check out the comments from Marcus on the composition of modern change management.
Second, given the company's age and history, changing established ways of thinking is also a major issue.
Third, customers expect Otis elevators and escalators to be safe, high-availability machines. The sheer size of the company's worldwide installed base creates ongoing pressure to ensure safety and reliability without screwups.
The company's digital transformation is a fascinating story that involves modern technologies such as sensors, internet of things, and even building interfaces between Amazon's Alexa and elevators.
Watch our conversation in the video embedded above and read the complete transcript. Below are edited excerpts from the transcript.
And, if you want to learn whether the "close door" button on elevators works, then read on!
As CIO, you develop systems that affect the product directly?
Marcus Galafassi: Yes. For example, the apps. We have done a lot of apps for our mechanics. These apps, they are apps that are linked to the customer experience.
That's the problem we have, the customer experience. We go into a service in a contract that we have with the customer. The customer doesn't see that I came there, and a frequent question was, "Did you come here and do what was supposed to be done? I didn't get any feedback."
We just launched a very simple digital tool using text as a basis. I'll tell you; the feedback we got was very simple. The feedback we got was great.
Again, when you do apps like these or also to our mechanics that can improve the right quality or can improve efficiency in the field like searching parts that I needed to order, this is a great opportunity where digital is beyond what I would call in a traditional company or the CIO boundary. We are doing things that go beyond our organization.
How does digital create opportunities to go beyond the traditional CIO role?
Marcus Galafassi: Sensor data is one thing that is addressing a lot of opportunities for us. In these two million units, from a contract standpoint that you have in the globe, we more than 300,000 connected today. We are harvesting this data from a sensors standpoint.
I think you talked about pushing the "close door" button, right? I bring the example of this "close door" push button. One of the major issues we do have in elevators is doors. Why? Because people are rushing in the day and try to arrive and then see the door closing. They try to hold the door. Then holding the door, again, affects the mechanisms we have, I mean the components we have in the door.
Sensors, we have put sensors in the doors, and you can understand that, throughout the whole data history, how much can you predict on this kind of sensor information, and how can we anticipate services to be maintained? Again, sensors providing this kind of data is very rich. If you look at doors, again as I said, 60 perceent of unexpected services are caused by this type of failure. If you can bridge the technology, the digital IoT capability along with the data analytics, it's a very powerful technology for us.
We have done more than that. In our deployment for these 33,000 techs, we have now 17,000 done. We have established a very strong network in terms of change management. We have more than 1,000 people that know how to support the technicians in the field. They know how to help them to sort the problems out because, again, mobile for me is easy but, from a technician that has worked 30 years in the industry, it's quite a challenge.
Again, that's an opportunity, I would call it. The technology has been evolved. I think the CIO has an opportunity and has to take this and move forward.
That's what we did two and a half years ago. We established a very nice strategy to transform and support our digital folks in the field. Then, of course, we're evolving for IoT moving forward and clean up the baseline, our backend in the company as well.
What were the business reasons for these technology changes?
Marcus Galafassi: We have done a lot of assessments in Europe, mainly because Europe is our mature market from a services standpoint. Again, most of the comments and what matters from a customer's standpoint is information. How can I make the information available to our customers, and how are they taking that information and, of course, helping them to drive?
I have a situation, for example. It's a true case where a customer told us, "Look; I have one person that walks every day in the morning, in the afternoon, in the night."
I ask, "Why? Why have this? To do what?"
"Just to walk in every level of my building and check if the elevator is running."
So, we have a customer that's telling you, "I have one person from my building walking every single day, every time, to check if the elevator is up and running."
Again, if you can connect the elevator, get the sensors, predict if it's going to be broken, or even more as we showed in the Shanghai Expo, we have what you call a customer view. We can show, in a dashboard, all the units that belong to the customer. Actually, the name of this is called the Campus View.
Instead of having the person walk and seeing if the elevator is down or is up and running, he can see the dashboard, and he can see it on his mobile, so think about it. Again, that's the way that you address customer experience. That's the way that you address the communication that was missing in the past. That's what the customers are looking for.
You want to simplify the customer's relationship to the elevator and its components?
Marcus Galafassi: We really appreciate technology. But, these 300,000 units, I told you some of them are still connected in an old-fashioned way. I would say ISDNs or corporate-based telephone lines. They're still getting the sensors data.
But, when you look at these technologies and you see the data around the technology, we can still apply, I would say, analytics on top. We can still apply data patterns around that, and you can still have information helping us and us helping the customer. Definitely, one of the major points that we have done in some of the apps that we have built around is to improve the right quality.
In the past, we used to have a technician come if a Motorola broke, as I call it, you know, then connect a cable and hooking the cable into the control board. OK. Now we have a wireless dongle, and I can connect even in the lobby without even touching the control board in the elevator machine area. That's one thing.
It improves safety as well, by the way. I don't need to have a technician downloading all the data associated with failures or potential failures. I don't need to touch the equipment. That's my point.
All of this is in the cloud. We started cloud four years ago here at UTC, our parent company. We are the pioneers of the cloud in our business. Again, I think that's the powerful information in the cloud at the hand of our technicians and at the hand of our customers. I think that's pretty much our vision.
What agile approaches do you use?
Marcus Galafassi: That's a great question. We started two years ago, Agile methodology. We learned a lot. It's not easy coming from a traditional waterfall into an Agile method. I mean the learning process, what it means to split into scrum teams. How can you put it all together?
I think the great news on this is, again, the technology has evolved a lot. I was a COBOL programmer. Don't ask me when I was. It was a long time ago. I remember I was sitting along with the other guy asking, "Can you do a report like that?" Then two weeks after, "Oh, by the way, I missed this."
Again, it was kind of, you come; you ask me what you want. I think I know what you want. Then I do something. After, you just realize it doesn't work.
Waterfall is over, from my point of view.
I think we can still use some waterfall technology or process to help us address, I'd call it, traditional IT systems. Some ERPs still require that. But, what we did, we actually have a very strong process with our parent company, UTC, and it is across all our business units. It's called ACE, achieving competitive excellence.
ACE has one of the tools called passport review boards. It sounds very bureaucratic, and sometimes it can be bureaucratic, as long as you want to. But, what we have done now, we have merged some of the key fundamentals and what is a PRB, as you call it here are, at the end of the day, milestones. I need to have a minimal design. I need to have a gateway. After the design is done, what's my build? After my build is done, what is my outcome that goes for production?
What did was merge the key gateways along with Agile to make sure that we would be fast with a sprint approach, creating scrum teams, but also controlling the quality throughout the process. These four apps, I think, work very well. We did eight apps in one year, pretty much. We deployed 17,000 phones, 17,000 mechanics, with an ecosystem using iPhone, and our MDM is AirWatch, pretty much at the same time, a year and a half. It was very, very aggressive, and we see the methodology for helping you, like DevOps or Agile. That's crucial.
If you want to go digital and you want to do it fast, you need to have, of course, some change management along with your customers, internal or external. Again, external, we are having a lot of experience bringing some key customers to attend some of the sessions and providing feedback in the design. I think that's the evolution.
Another important thing, we started this pretty much alone. A year ago, UTC, UTC Digital -- UTXD, as we call it -- has launched an accelerator in Brooklyn where some of these capacities were not present there at all. We started, again, an hour away from a notice elevator standpoint, but the capacities like design thinking, ideation, incubation, product management, not project management, product management and, of course, analytics. We are creating these capacities inside the house. You can go and show off the results outside, but I don't think it's healthy.
If you want to just really turn your company digital, you need to have this kind of capacity. You need to have a design and thinking process. You need to have Agile, DevOps. You need to have product managers taking care about the product evolution, the digital product evolution throughout the period. And, of course, in AI, analytics are components that can help you improve and enhance the product. Again, it's a mix of methodology, capabilities that you need to have in-house, and you need to foster this more and more and more in our IT organization.
How do you change an organization that is 165-years old and so well established?
Marcus Galafassi: It's, not all about technology. Technology is great, but it's technology; it's people; it's process.
I think, what you have done, we addressed a very strong champions network to help us achieve goals. I'm talking to the apps aspect now. We have 1,000 champions across the world. I think these people have been trained. These people have been fundamental to help us turn this. I call it adoption because it's really adoption.
You can give a very nice phone to a technician, and he can YouTube. He can do this; he can do that. But, at the end of the day, you want them to use the apps and make his life more efficient, the customer's more communicated, and the experience better.
The key fundamental point is if you go to this technology and if it's 165-year-old company or if it's a six-month old company, it has to have a change management concept in place because, without that, the technology itself is not going to help you. People have to adopt. If people don't adapt and adopt, then you have a problem.
I've got another example that our service leader gave me, and I was very happy about it because we deployed along with these 1,000 champions in our change management network. A lot of communicates decided on our Yammer setup.
There was a guy. I don't remember his name. A mechanic, he was having a problem. It's like China time zone back to California time zone. The guy was typing in Yammer. He said, "I have a problem, specific problem with this setup in this machine with this control board," and so on and so forth. A guy overseas, he answered and said, "Look. I had the same problem. By the way, I did this."
Think about the power of this. This is a true change management. [Laughter.] It's a tool that is across the board. People are talking to each other virtually. They don't know each other face-to-face. They never met each other face-to-face. But again, that's, for me, the power of the technology. But, we had to create this network because, if we don't create the network, if we don't create the right groups, we just can create, as they call it, a little bit of confusion, you know. We need to use the right technology to help and evolve the newsletter to type it once in a true, digital communication.
Modern change management means collaboration?
Marcus Galafassi: Yes. Pure collaboration.
Pure collaboration and you have done some tooling as well, some tools where our mechanics if they have a problem, can call an expert. They can Skype themselves. They can show on the phone where the problem is, and someone is going to help them to address the problem.
Before it was someone trying to call someone. We are digitizing all of our technical information, which is the basis for searching parts and everything else. In the past, we used to have binders in the trunk, and you go back to the trunk in the car. You open it up and see what's the part I need.
Definitely, technology helps change management, but we need to organize that and structure that properly and have the right champion network, have the right expectations how to use the technology, what can be done, what cannot be done, [and] create the collaboration. Collaboration is key. I think the technology helps, but if you don't have structure, then it becomes another app that is not going to be used.
What metrics do you use to evaluate digital strategy?
Marcus Galafassi: You have to address adoption. Let's talk about adoption metrics.
We have adoption metrics in terms of usage of the apps. We have a process also to collect, in our incubator, ideas that ties, again, with the KPIs that we have established.
I think one of the main concerns, along with the change management, is how can you drive adoption? Adoption, again, is relative. It depends on the person's experience, the person's expectation, but we have set up very nice KPIs around how the apps and the usage of the apps have been, on a monthly basis, collected.
We have a process in place. We have improved it with a lot of design thinking and improved the UX, the experience. We are trying to combine some of the apps to have the right business flow or process flow from a mechanic's standpoint.
We needed to map personas. We needed to map our ecosystem where these personas are going to use these and for what purpose. Usually, I mean traditionally, we have not done it in the past. So, I need to put myself in the shoes of a mechanic, and I need to in a hoistway, which has no cellular communication at all.
When you build up an app and think, "Oh, an app is great," oh, but there is no offline capability. Okay, I think it's a big mistake, right? You are in a building, in a hoistway, completely closed, and there is no network coverage. You cannot build up an app thinking that the app is great if there is no offline capability.
We need to have, again, the persona. You need to put yourself in their shoes and have the right empathy to make the right process, the right tool, the right app for them.
When you design a product, and I talk about an elevator, we need to think in that perspective as well. How can we integrate the sensors? How does that sensor go into the cloud? What are the AI opportunities we have around that? Can I put Alexa to call the elevator? Yes, of course, you can, and we did it, by the way. You're more than welcome to come here and visit us in our Bristol Tower.
Bristol Tower is in Connecticut, in Bristol, and it is the highest tower in North America. We have a prototype of that we show to customers. They love that. Again, these kinds of ideas, that's what you want as a spirit, as a digital company.
How are you using the Internet of Things and why?
Marcus Galafassi: I mentioned before the Campus View, which is one of the customer dashboards that we have presented in the Shanghai Expo. The Campus View gives you how healthy the customer fleet, the elevators, are in their respective building.
You can see the available red and green dots. Why it's red is because we can have an elevator or the unit down, shut down. Also, you can see, and you click. It can start to turn green to yellow, yellow to red, and you can move the time. The time means I can move ahead a little bit and see in three months that yellow becomes red. That's what we can see. Of course, the customer has the information available.
We are using sensor data. We are feeding off this data in our cloud. We have done a condition-based maintenance to predict using, again, doors as 60 percent of our call-backs, as you call it, or unexpected visits are driven by doors problems. We have done very well, along with engineering and our research center here, prediction in how these patterns of failures could happen and then how they would happen. We can anticipate that failure. That's what addresses not only deficiency in our site but, also, we prevent to have a unit in a shut-down mode, which is the worst thing that can happen to our customers. Customers need our elevators 100 percent of the time up and running. That's pretty much the whole business value we see with IoT in our case.
Are the close door buttons on elevators real or fake?
Marcus Galafassi: If the customers ask you to shut down, we're going to shut down the button. [Laughter.] That's a customer request.
[When people force open an elevator door that is closing,] the whole mechanism [of the closing door] (sic) have to move back. If you have extra strength, it's even worse. This completely interrupts the momentum of closing the door, so we don't like it [and can cause the doors to require service].
It can be perceived as a fake button. But, I let you judge that. [Laughter.]
CXOTalk offers in-depth conversations with the world's top innovators. Be sure to watch our many episodes!