Martin Eigner talks The Industrial Internet & Industrie 4.0

Maybe you have listened to my speech, so I have to say Industry 4.0 is mostly reduced to production and not for the full value chain. So I think Industry 4.0 is a typical German word. Let's talk about the American term Industrial Internet. So because we have to, when we talk about value chain, we have to talk about product development, production planning, production, and especially service. And I'm quite sure that we will improve the value chain much more in the area of product development and from the beginning and at the final end with new service-oriented business models. In Germany or Central Europe, we have a real high level of degree of automation in industry and in production. So for me, product development, new products which are able to communicate, and services which are based on communicating products, that's the real revolution of Industry 4.0, Industrial Internet. The main investment, I think we are talking about a lot of IT tools in the process chain. So we have the— in German, we differentiate in 4 levels. We have authoring tools, we have team data management, we have what we call the PLM or engineering backbone, and we have the MRP system on top. I think the— for the Industry 4.0, we have exactly the same investment in processes and tools as for mechatronic. So there's no big difference from the toolchain between mechatronic and Industry 4.0, Industrial Internet. We have— we will have more products with software, electronic, and mechanic. What is new is the service-oriented business model. So we need to invest in two things. One, I think we have to come to a platform, a PLM platform, which is able to support interdisciplinary work. That does mean they have to integrate all this, all the development of mechanic, electronic, and software. We have to invest in the early phase, what we call model-based systems engineering. So we need, and when we start the product development, we have an interdisciplinary phase which is not discussing about any discipline, and that is what we call the modeling and specification based on SysML. That's new, but it's an authoring system like any other authoring system. And then we have the simulation based on Modelica, Simulink, and System C. So we have new tools, but these tools are only to be used in mechatronics. The only difference is indeed we need models to describe services. So because when we have products which related service, for example, autonomous parking or kitchen aids which are controlled by smartphone or communicating each other, then you build up a service or predicted maintenance. That is a service. And as an OEM or a solution provider, we'd like to to sell the service together with the product. And so the service must be embedded in the product development. And that is a difference of industrial internet. We have a totally new discipline. That does mean service-oriented development and service providing in after-sales area. And we have no methods right now to describe, no formula methods to describe the business models. But otherwise, all the things which are valid for mechatronics are valid for industrial internet. It's not a big revolution, so maybe it's an evolution. Upstream, I mentioned before upstream, I think the typical approach of a PLM system starts with the engineering bill of material, that's standard. And so we administrate the mechanical CAD system, hopefully the electronic CAD system, and sometimes even software is embedded into the engineering bill of material. I think that is a starting point of a PDM environment. I wouldn't like to talk about PLM, and the upstream integration should be what I mentioned into the model-based systems engineering areas. That doesn't mean we have to administer requirements, and not only the textual from doors, we have to transfer the requirements into this tool system manipulation language or a modeling language, SysML, and then we have to buildup and functional breakdown and the logical breakdown. For electronic and software, we need behavior. So, we are right now discussing model-based systems as a totally new approach of product development. We have an interdisciplinary area based on this MEL, it's an UML dialect, and we describe requirements, function, logic, behavior, and then we transfer to physics. Because all these elements are connected, we know which functions are realized by or driven by a requirement, which logic is belonging to functions, And so we have a wonderful network on a non-discipline-oriented level, and then we can talk about to configure it down to the, to the discipline. So upstream: requirements, function, behavior, logic. Very easy. And of course, early simulation based on Simulink and Modelica. Downstream is a little bit, is a little bit manifold. So Even now we have not a very good integration to simulation. I think, see, when you remember the V-model, we have the requirements, we go more in detail, then on the other side we have the physical or logical or virtual models, and we are testing always against the requirements. But that does mean we have to simulate or to test. We have to virtual simulate or we have to physical, we have physical tests, and all the simulation results and and our test results must be compared or validated against the requirements. So that is indeed, it is a circle. So what we have to do is more improvement in integration of simulation into the PLM environment, because I need to, when I have the requirements in, I would like to have the test results or simulation results. That's one topic. The other topic is when we talk about service-oriented business models, I would like to integrate I would like to use PLM or extended PLM as a service lifecycle management platform to get all the data back from the market, from the product. Good question, because I think when you start the development process, we have to handle, we are faced with very creative engineers. And says there is a mismatch between creativity and administration. I think that's a matter of fact. And as more people are creative, as more they try to refuse any administration, and that is a real problem. So we have to build up a balance in the very early phase where the people are working fully independent of any administration. They are in a creative silo, they develop new products, new ideas, real innovations. But at a specific point, let's talk about a specific gate where some communication transfers the barrier or the barrier between departments or even the company. So we have to inform a supplier or customer, and so you interchange any communication, any information, and that's from my point of view where PLM has to start. Here we have to We start with an administration of function and requirements and all this stuff. And as just at the moment we start to administer it in the PLM system, we have to agree that we have to use the typical engine change management from PLM or MRP environments even for the new object classes like requirements, function, and logic. So we have to keep in mind that the change process for mechanical parts or for hardware is different to software, and we have to accept that the change management for model-based system elements are even different too. So that is something that we have— as a matter of fact, we have to trace all the information based on ISO standards. There must be a traceability from requirements to even to recycling, and the only way to do it is to integrate all these new objects into the PLM system or an administration environment and to keep track based on configuration, on change management, baselining, and that's the end of configuration management. It must be configured. The question is general. What is the efficiency or the impact from PLM into the full enterprise business process? Is PLM really improving the value chain? Question mark. I'm not so, I'm not quite convinced because most of the so-called PLM installations are really PDM, department-oriented PDM installations. I'm quite sure at the end we will have a common value when we have full integration, but anyway, we have to keep in mind PLM is administration, so we are not able to substitute creativity by PLM. So we can support creativity and we can administrate the results of creativity, and I'm quite sure that requirements from the market or from the legal point of view, like traceability in case of accidents, product reliability, must be supported by a PLM solution. And as more the products become complex, as more different disciplines and services are belonging to the product, as more complex is the configuration management or the change management along the product development and product lifecycle process. Good question. What I mentioned, for the hardware, industrial internet is nothing new. Everything we knew I think 20 years ago I have it in Bang Olufsen and with the remote control I was able to switch on the light or I can dim the light. So was it industrial internet? It was based on Bluetooth or infrared, I don't know. Well, I'm not so worried about how the products or the components are communicating, whether it's based on optical control, RFID, or Bluetooth. Or even the internet. I think that is a technical level. I think they are communicating, and once again, these products are based on mechanical, electronic hardware and more and more software, and there's no difference from mechatronic. All the— or most of the requirements related to IoT, iOS, or industrial internet are related to build up an environment we can handle mechatronic. We have to keep in mind that the It's the actual way to handle mechatronic parts. Mechanic starts, then there's a big barrier or gap, and this was a mechanic development to the electronic. They work something, and then at the end, the software engineer has to finalize the solutions. It is not what we call the synchronous engineering. It's sequential engineering. And once again, the only difference in IoT or industrial internet is we have to take care about services. And we will have maybe an industrial equipment manufacturer, he will add the service to the product like predictive maintenance. Then these services delivered by the, by the OEM. Maybe there's a quite different solution. You have autonomous parking, you have a parking slot, a smartphone, and a car, and you would like to have to leave your car in front of the parking house and The car is autonomous driving into the parking slot. Who is the stakeholder? Who is the provider from this business model? That's not clear, but that must be described. It's a different type of collaboration, but it's not the huge impact to PLM. All the things what I talked about, the model-based systems, engineering service orientation, I think it's valid from electronic to That is dramatically so. I think especially the investment goods, they haven't condition monitoring right now. They have some sensors, they are wired, we call it, they have a wired communication. So condition monitoring is not defined parallel to industrial internet. It was, it is 20, 25 years old. The European manufacturing and assembly industry handled more condition monitoring. What is new right now is that we have many, many different sensors, many, many different information getting from the machine, so we are able to control all the state of the machine, so we can predict failure or any stop of the machine, that something is predictable, or if the machine is not working, we know which component is broken and we can deliver the spare parts inside a very short time. I think that is a business model. Predictive maintenance is very close to the condition monitoring, what's going on. It's a better method, it's more elegant, it's faster, and we have more information. What's totally new is for consumer goods, we would have totally new approaches. So think about— that's a joke we did with my students— we thought about the demographic development of the people in Germany. We have— normally it looks like a tree. We have some old people, many young people. In Germany, it's vice versa. We have more old people than young people. And so we have an— we need an industry and we need help for elder people. And so we, for example, very easy product would be a rollator based— very nice— based on— it looks like a golf caddy. The golf caddy has an iPad in the middle. The iPad, there's a big red button, "Bring me home." So maybe I'm lost in space in a town and I would like to go home because I become old, maybe I become demented. I have to hit this button. The rollator brings me home. Or the next point is my pacemaker is related to the iPad. If the frequency is more than 180, there's a signal via internet to my iPad. That, oh, Vonnieb Aigner has high blood pressure. The next medical care center will be informed. Medical care center looks to the insurance, whether I'm private insured. If yes, ambulance is coming very fast. So it is more or less a joke, but it shows that we have totally new possibilities in smart health, smart buildings, smart energy, smart grids, smart mobility. And that is the The impact and what we have to rethink, we have to be more creative, more innovative to use the possibilities of communicating hardware. And once again, the revolution is the service-oriented business model. We will find out that there will be a huge market for the future.