Thank you very much.

Thanks for the introduction.

It's really a pleasure to be here and give this talk on cyberphysical systems.

So what I'm going to talk about is really joint work, as you might see, with several

people, both from Munich as well as from MIT as well as from a lot of other universities.

Okay, so I guess one question that might be of interest to all of you is what really are

cyberphysical systems?

This is a term that all of us have been hearing quite a lot since the last couple of years.

So I would not actually try to define cyberphysical systems because this is more complex.

So there are actually many, many interpretations or examples.

The first is network embedded systems.

Often these are referred to as cyberphysical systems.

Lot of people also refer to sensor network based systems.

Another term is also internet of things, commonly also used in Munich.

Also embedded control systems, and embedded control systems is kind of the definition

that I would assume for the purpose of this talk.

But what runs through all of these examples is that these are systems that where a computational

part closely interacts with a physical part, right?

So as you might guess, lots of examples like an air conditioner which has a control software

in it and tries to interact with the room and then maintain a certain temperature.

That would be, for example, a cyberphysical system.

But however, given that we are not too far away from Audi and also I come from the city

of BMW, we will actually use automotive as one of our running examples.

You also heard examples of automotive from the talk of Professor Wilhelm also, Professor

Glass gave this example of automotive.

So let's look at automotive once again.

So if you look at what has been happening in the domain of automotive over the last

15 to 20 years, we can see that this is the first picture on the top.

It essentially shows that most of the innovation today is no longer in the mechanical engineering

aspects of a car.

So like let's say 10 years ago, a car was primarily something of a subject that was

studied by the mechanical engineers.

But today at least the good news for us computer scientists and electrical engineers is that

most of the innovation is currently in the software and the electronics because the mechanical

aspects have more or less matured.

And the figure below shows you that the way, for example, memory and the number of ECUs

have been going on increasing.

And currently we have high-end cars that have almost 100 or even more electronic control

units.

Each of these ECUs have one to two processors.

So we are essentially talking about something like 100 to 200 processors inside the car.

And this is a computer that has one processor.

So it's almost like 100 to 200 such computers built inside the car.

And the question is, what are all of these processors doing?

So they are running several distributed control applications, applications coming from different

domains like safety-critical applications related to comfort, applications related to

driver assistance systems.

So this would be then a view of a modern car with a lot of processors that are, of course,

connected in some way through even kilometers long cables and running a lot of software

related to different control applications.