Thought Leaders in Artificial Intelligence: Michel Morvan, Co-Founder of Cosmo Tech (Part 4)

Sramana Mitra: Let me ask you how. There are certain core technical decisions or choices you have made here to enable this kind of decision making. Could you give us some insights into what those are?

Michel Morvan: I mentioned the fact that we created a modelling and simulation platform that is dedicated to the modelling and simulation of very complex systems. What do we mean by that? Systems like the one I just described in which you have different subsystems, each subsystem alone cannot explain the global behavior. In a complex system, sometimes they do not evolve at the same pace.

We can model electric transformers globally, but we cannot model certain parts of it. This is what we call a complex system. If you want to be able to model that, you need to be able to model with different kind of models. Maybe when you’re going to talk to experts, they will say, “The best way to think of that is to think in terms of partial differential equations.” Then, you need to model it following partial differential equations.

For another subsystem, another expert will say, “I want to model it with a multi-agent system.” At the end of the day, you need a platform that allows you to put all these different models that are completely heterogeneous and are all interacting with each other. Sometimes you can add other models like a neural network model. You need to be able to put that together and be able to have a simulation engine based on that. That’s one of the big subjects addressing complex systems.

This is something that I started when I was a university professor. I was leading a research institute on complex systems. At that time, we were modelling living systems. As you can imagine, living systems can be very complex to model. Because we wanted to model at a different scale, we had to create a language on the modelling platform that would allow you to model all the heterogeneity that you can have in a living system.

When I was leading this institute, we created the modelling and simulation platform that was not only dedicated to living systems but was quite generic. After a couple of years of using it in academia, we created a startup. The startup used the IP and bought it. We started to develop different solutions, which were always on the ability to model this complex system and to simulate them.

Sramana Mitra: I still don’t have the nuggets of what drives your decision-making capacity. I understand you’ve created all sorts of systems for modelling this and that. Can you give me a simple crystallization of what drives the kind of stuff that you are able to do? What are some of the algorithms that drive this?

Michel Morvan: You need to be able to model very different systems. Let me give you another example to illustrate what I mean. This is a subject we’ve worked on in the past. Assume you want to model a city. You want to model this city because you want to know what is the best decision you can make for reducing air pollution.

If you want to model it, there is not a unique way. You need to model the cars and the way people use their cars to go from home to work. You have to model the buildings. You also have to model the air and the pollution going into the air from the cars and the way this can move with wind.

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