6 May 2020

Let’s take AI to the next industrial level, starting with Automotive!

Artificial intelligence – AI. Yes, the topic is almost passé. After all, brilliant minds have been working on it for over 50 years and it generated many hypes. But it was always the same old problem: there was not enough computing power. Thanks to today’s modern computing capacities, AI is now “superior” to people in many areas.

From Data to Value

So, although the topic is not new, there are good reasons for this article especially when you consider that we at Siemens approach the subject from a slightly different angle. We ask ourselves where the data that’s fed to the AI algorithms comes from and how we can industrialize the use of AI. Fortunately, we also have the answers to these questions because in our role as a leading provider in automation and industrial software and especially as integrator and turnkey solution provider in Automotive we understand how production processes work. In other words, we know how to get from data to value. For many AI applications – in the automotive industry, for example – you also need computing capacity on the shop floor. We can supply the necessary industrial hardware and software.

Maximize your return on investment

Another advantage for customers is, to collaborate with us on “AI field deployment”, which I consider one of our greatest strengths. This involves rolling out developed AI applications in additional machines and factories. Only through an effective replication, the ROI can be maximized. I’ll give you an example. Imagine the following: In your production, welding spots have to be set on PCBs (Printed Circuit Boards) or other objects, like car parts. The quality of these welding spots is critical for productivity and the quality of the product. Using historical data that has been collected, you can now train an AI application to distinguish between “good” and “bad” welding spots – in a fully automated way. This not only improves the product quality but also increases throughput because inspection is automated. So far, so good. But to realize a quick ROI, this AI application would now have to be implemented at all the other locations, preferably with a push on a button.

In addition, not all manufacturers have the necessary IT resources to customize and implement AI applications. And this is where we come in, with our Industrial Edge portfolio and our global automotive team. We help you to bring the AI application to all your factories so that you can achieve a fast ROI. This can be done with the Industrial Edge Management System with hosting either on-premise or in the cloud, depending on your requirements. It’s then possible to remotely install, manage, and monitor all applications and connected edge devices centrally and from anywhere in the world.

Co-creation – achieving more together

Speaking of help, not all companies in the automotive industry have their own army of AI specialists like the large corporations do. To help these customers, we use co-creation and IP sharing. How does it work? Let’s return to the example above. Let’s assume we’re not only in charge of deployment, but we’re also working with the customer to develop the entire AI solution via co-creation. Naturally, the development partner also has exclusive rights of the new application, in terms of both location and time. But now we also make this AI application accessible to other manufacturers, who don’t have their own development department, via IP sharing. This makes AI affordable for smaller companies as well.

So, we deliberately position ourselves as a partner for application development and rollout. But, when we develop applications ourselves, we prefer scalable solutions that enable us to meet the requirements of one or more customers.

That’s enough about Siemens and AI for now. In a follow-up article, I’ll address some specific use cases, among other things. Meanwhile, I look forward to your feedback.

Related Tags