Thought Leaders in Artificial Intelligence: Michel Morvan, Co-Founder of Cosmo Tech (Part 3)

Michel Morvan: The CEO has to decide to delay these investments and to use the money for some maintenance for some other equipment. To solve it, they have to make choices. This is the first problem. At the same time, you have the renewable energy entering the system. This changes the way some types of equipment are aging. You have to take that into account when you make your decision on what to delay and what not to delay.

These happen in the context where a great part of all the employees are going to retire in the next 10 years. If you change your policy for maintenance or renewal, you will need different employees having different skills. You have to put in place an HR policy that will take that into account. If you don’t do that, you will have a plan to maintain and nobody will be able to do it.

As you know, an electric network can burn a forest if there are problems. Just for one second, you think of this. You have this extremely big network that has a lot of spending every year. Now you will have to change the policies you use for maintenance and renewal. We are talking about big money. We are talking about €700 million every year. Assume that you are the CEO of a utility company facing this problem. What would you do to decide which equipment you can delay the maintenance for?

You’re going to ask the asset managers and say, “I know that an electric transformer is €20 million. I know that you need 18 months to produce it. Don’t you think that we could delay?” Of course, the asset manager would say no and say, “You should ask the people managing the circuit breakers.” Then you go to the people managing circuit breakers and they’ll say, “Circuit breakers are so important to the network.” Everybody is going to say that. Your main problem is that you don’t know how to choose.

If you want to be able to make a decision, you need to know what is happening when you have failure in one equipment. You have a circuit breaker that breaks or a transformer that doesn’t work. You have to understand what is going to happen when this takes place. Let’s say, for example, a transformer breaks. You send someone to observe. You find people with the right skill. During this time, the current doesn’t go through the transformer because it’s down.

You can say, “At the same time, I can use the time to maintain some equipment which are related to this transformer.” You can ask other people to come and maintain that. At the same time, the current is going to go through another line because this one is down. This is going to overload some other equipment. It’s a big risk that has a cascading effect. We encapsulate the expertise of all experts. What happens when a circuit breaker breaks? What happens when some other thing breaks?

We have a modeling and simulation platform in which we can encapsulate the dynamics of each part of the system coming from each expert. We talk to the electric transformer expert and encapsulate his knowledge into a dynamic model that is able to describe the causalities. We did that for all kinds of assets. We also encapsulated the expertise of the HR people.

Since we have this unique modeling and relation platform, we were able to combine all these different models. What happens here has an impact there. Human resources are related to assets and everything is related to everything. Because we have this unique modeling and simulation platform that is dedicated to the modeling and simulation of very complex systems, we had the ability to encapsulate this expertise. This is the first level of augmentation of intelligence. We put it together in the same model. This was the first step.

The second step is, for some equipment, there is knowledge that is not known by the experts. For example, the exact aging curve of some equipment that depends on some past things. Because there’s a lot of data from the past, we put machine learning algorithms that know what is the expected lifetime of an equipment. Once we have that, we created a software that is used to optimize the decisions and that runs simulations.

If the CEO says, “I think I can do something. I can delay the replacement of these transformers by five years.” We simulate that and we can say, “Well, the idea is good for the first five years, but there is a very high probability for you to have a major outage in 10 or 15 years.” Then you can test different strategies. This is a very complex example. This software is used by different electric utility companies to optimize the maintenance and renewal of their equipment.

This segment is part 3 in the series : Thought Leaders in Artificial Intelligence: Michel Morvan, Co-Founder of Cosmo Tech
1 2 3 4 5