Next Generation Manufacturing: Production Systems that Think

Next generation manufacturing is all about production systems that can think for themselves, much like self-driving cars…

Do you remember how you felt when you first heard about Google’s self-driving car? If you’re like us your reactions probably evolved from: “Interesting,” to “Well, if anyone can do it, it’s Google,” to “This is really going to happen, and probably in my lifetime.”

The company, which is best known for a product that has nothing to do with cars, transformed the collective thinking about cars and their potential. Google is attempting to have a similar effect on other industries, such as networking and medicine. Google allows its employees to work on “audacious” projects, which seems to be a core element of the company’s philosophy.

Siemens is equally as innovative in terms of its audacious vision of next generation manufacturing. We are making significant investments in our own production capabilities – digital manufacturing software, automation equipment, and communication protocols – to learn how to manufacture smarter and lead our customers into the future of manufacturing.

Ultimately, in the not-too-distant future we see this: a manufacturer receives a digital model of a new product, and based on the information in the model, the production environment configures itself to produce that product.

Some refer to this scenario as “self-organizing manufacturing processes for highly customizable products.” It represents a global revival of the manufacturing industry, and is being driven by government funds, market forces, and technology megatrends. In Germany this endeavor is called Industrie 4.0. There is also the Smart Manufacturing Leadership Coalition in the US.

Why do industries need this? And why would governments sponsor it? From industry’s point of view, next generation manufacturing offers a way to meet customer demands for new, high-quality customized offerings at ever-shorter time intervals. It also has the potential to reduce resource utilization, which will help manufacturers cope with growing cost pressures.

From a governmental perspective, one driving factor is the fact that people in emerging markets still need many things. A recent article in Time, GE Makes a Big Bet on Manufacturing, called this a megatrend and described it this way: “. . . emerging-market economies are entering a period very much like the post-World War II period in the US.  Countries need houses, bridges, roads, airports, and all types of consumer goods in unprecedented quantities.”

Countries such as the US and Germany would like to have those products manufactured within their borders, so that their workers and their economies benefit from some portion of the $20 trillion per year that McKinsey Global Institute estimates will be spent in this way by 2025. Factories capable of autonomous production will be able to quickly adapt products to the wide range of consumer preferences found in the emerging-market nations.

The Internet of Things

In this article, we will refer to the move toward self-organizing manufacturing processes more simply as “autonomous production.”  A key technology megatrend driving autonomous production is the Internet of Things (IoT).

In popular parlance, the IoT refers to “things” – such as Nest thermostats, glucose-monitoring devices, cars with collision-detecting sensors, even pets with embedded microchips – that relay information about themselves via the Internet to someone, or more accurately, to some computer, that uses the information in an intelligent way.

The IoT can dramatically change the status quo. People with diabetes, for example, are much less likely to go into a coma if they can receive a notification on their smartphone that their glucose is low.  Losing a pet no longer means printing and distributing flyers and hoping for the best. You can immediately see the animal’s location on your smartphone.

The IoT is not limited to consumer products. The aviation industry, for instance, is undergoing an IoT-based revolution by harnessing information generated by engine sensors. Engine manufactures now have access to enormous amounts of data about engines in flight, which they are using to find ways to reduce fuel consumption and become aware of anomalies while a flight is in progress rather than after the fact. This is changing entire business models, with Rolls Royce and GE now contracting for hours of operation, not sales of engines.

Implications of IoT for autonomous production

The IoT has been made possible by massive technological breakthroughs in computing power, miniaturization of wireless sensors, high-capacity networks, and big data analytics. Another important element is the fact that each of these technologies has come down so far in price, especially with the availability of cloud computing, that its use can spread widely.

All of these technologies – computing power, miniaturized wireless sensors, high-capacity networks, and big data analytics – are already in use to some extent in manufacturing. So it makes sense that we are beginning to see an industrial application of IoT. In fact, there’s now a term for it – the Industrial Internet of Things (IIoT).

The deployment of the IIoT in manufacturing plants is making it possible to extract much more information about the production process than has previously been possible, even though manufacturers have already been collecting a great deal of production data. The more extensive dataset made possible by the IIoT will ultimately move us toward autonomous production by making it possible to quickly update a production model to adapt to changing conditions, such as a custom order.

Compare this with the current practice in which a production system is designed and optimized to execute the exact same process over and over again. In an autonomous production scenario, manufacturing systems will have the flexibility to adjust and optimize for each run of the task.

Let’s use robotic operations as an example. Today robots are programmed to perform specific tasks. As long as certain, pre-defined surrounding conditions are met the robot will always execute the task identically. In a future with autonomous production, the robots will receive a task and will have to determine how to perform it in the most optimized way. Theoretically, for each run of the task, the work may be done differently.

Kineo: Autonomous Robot Navigation

This will also hold true for a complete system. For instance, in the framing station of an automotive assembly plant, where major parts come together to form the complete body, each major component will be measured and the system will adjust itself to make the mating between the components optimal. Or each door panel opening will be measured and the best matching panel will be selected to the specific body (vs. the current just-in-sequence method).

Where are we now?

Even with the growing proliferation of the IIoT, we are not at the point where autonomous production is widespread or routine, although there are some plants implementing some elements of it today. What is already widespread and routine, however, is the solid foundation for autonomous production, which can be seen in practice in the many plants that have adopted digital manufacturing.

Digital manufacturing solutions provide manufacturers with vast capabilities for designing and evaluating their processes virtually. In a digital manufacturing environment such as our company’s manufacturing planning and management solutions, the physical world is replicated in a model-driven database. Digital tools and methods are used to design the physical manufacturing system, including its logical controls.

The result is a comprehensive virtual model of the manufacturing process that crosses multiple engineering disciplines such as tooling, process, logistics, quality and product. Digital simulation tools make it possible to validate and optimize the processes, tools and the control algorithms, and the interactions between them, all in a virtual environment prior to commissioning the system on the shop floor.

Beyond digital manufacturing is the digital factory, which is comprised of additional technological layers. A digital factory requires an infrastructure to connect the devices, the ability to identify where connectivity legitimately adds value and is not merely intrusive, and software platforms that will unlock the torrent of data.

Siemens is delivering big parts of these concepts today with its Digital Factory portfolio. The solutions included here – seamlessly integrated hardware, software and services – are already enhancing the flexibility and efficiency of many company’s manufacturing processes.

To put the Digital Factory in its proper place on the road to autonomous production, let’s look first at how it can improve manufacturing flexibility, since that is a critical element of autonomous production. Traditional production uses a sequential process flow between production modules (a moving line), where each module has a dedicated task to perform in a given sequence.

A flexible process flow between production modules, made possible by Digital Factory solutions, allows different modules to be configured for each production instance. Such a production system is more resilient to changes, and allows greater variation in manufacturing scenarios (production mix, production volume) and product range.

Let’s also look at how Siemens Digital Factory solutions improve efficiency.Digital Factory solutions optimize production asset utilization by constantly monitoring, controlling and analyzing the production ecosystem and making online decisions. They can do this for physical assets, such as machines, inventory or energy consumption, as well as for more amorphic assets, such as time to completion.

 The next steps are already being taken

 Additional technologies are still needed before autonomous production is a reality, but they are arriving on a regular basis. For example, Siemens PLM Software and our partner, Bentley Systems, recently introduced a new point cloud technology that makes it possible to capture the exact position of the production and logistical assets on a shop floor, providing in almost real-time the actual conditions of the shop floor.



In this past, this task took weeks, with many engineers needed to scan and measure the production facility. By eliminating that step, the new technology will permit faster modifications, a critical requirement for autonomous production.

Similarly event-driven simulation, (also known as discrete event simulation) which is already available, will play a bigger role as a fundamental tool for enabling autonomous production. This is because behind the flexibility and autonomy of this type of production scenario, there are very rigid rules that the system must adhere to. The self-driving vehicle is a good example – without strict rules, these cars would wreak havoc. The challenge is to move from nominal planning to variable planning, where the result is driven by the varying surrounding conditions in the ecosystem.

We already know that manufacturing is a key driver for economic growth, attracting investments, spurring innovation, and creating high-value jobs. All of the breakthrough and developments that are happening now – the building blocks for autonomous production – are already adding economic value. Imagine what will happen when autonomous production is routine.

Not only is it likely to be the revival that many predict for the manufacturing industry. We also see autonomous production as a way to address many global challenges such as a growing and aging population, climate change and resource scarcity.

It’s an exciting time indeed and Siemens is proud to be a leading force driving technological advancements to make autonomous production a reality. Keep watching Siemens as we deliver the results of our own “audacious” projects in the years to come.

To learn more, download our “PLM for Manufacturing” white paper about how to design and validate processes in a digital manufacturing environment.