The use of AI in manufacturing is transforming the industry. How do AI technology and data insights fit into this trend? Vice President for SAS in the Nordics Keld Zornig sat down with Pasi Helenius, Manager for Customer Advisory Commercial and Manufacturing Nordics, for a talk on the topic of data-driven innovation.

 

Data-driven innovation

Keld: To start, how do manufacturers use analytics and AI to drive innovation? Where does innovation begin?

Pasi: First of all, executive support is required to jump-start the process. Change really starts to happen when business goals are shared companywide, from top to bottom and from bottom to top: When it is not divided by departments or disciplines. A single department or factory is too small to drive change and innovation alone. It takes the entire team to work towards the company’s innovation goals.

Keld: So, in essence, if you run an AI project in manufacturing, it needs to be done in close alignment between HQ and all factories. When this has been orchestrated, how do companies decide where to implement automation? Does it begin with the technology available?

Pasi: No, we really recommend that companies start in a different place. They need to find, prioritize and implement use cases. When you start with a problem, it is a lot easier to find the right solution. One example of how to start with the problem is the transformation in steelmaking at SSAB. SSAB had a goal to have a fossil-free value chain from customers to end-users. They needed to use advanced analytics and IoT to find ways they could digitally transform their current processes and become more sustainable.

The biggest challenge

Keld: That is a great example of starting with the vision and then looking at technology. But once you get there – today, there are many comprehensive capabilities in technology, including the ability to deploy on edge, cloud, etc. But automation within manufacturing seems to still have untapped potential. What are the biggest challenges when it comes to adding AI to the manufacturing processes?

Pasi: One of the biggest challenges lies with us people as people! Change is difficult, especially if we do not understand “the why." Why is automation the answer? How will it improve the processes people are already doing? What is the goal? Operators need to understand their role in the overall company goal. How do the process operators need to change their way of working? What are the skills required to implement these new automated processes? Will management offer courses if required? Automation and AI tools are there to help free up their time to focus on more detailed work.

Don't get stuck on the basics

Keld: That is a key point: AI and automation are tools, not replacements for people. Creating a common understanding of this is crucial for people’s motivation to participate. If manufacturers are just starting to add automation tools to their processes, how can they avoid the pitfall of getting stuck on building the basics, such as data lake, connectivity, etc.?

Pasi: It is essential to think about value or competitive advantage. It is no longer a nice-to-have but a must-have. Manufacturers need to simultaneously do end-to-end analytics use cases while building out the basic capabilities. Otherwise, you will have too big a project to start with business value too many years ahead.

It is essential to think about value or competitive advantage. It is no longer a nice-to-have but a must-have.

Keld: I fully agree. It is about moving from vision to concept to value in rapid succession. If you are unsure how to scope this, I assume a call to your customer advisory team will be a very good starting point, Pasi?

Pasi: It most certainly will. Our team has extensive experience in scoping data-driven innovation projects in a number of industries, not least manufacturing.

To learn more about the future of manufacturing, visit our industry hub page here.