Solving the Industry 4.0-puzzle in discrete manufacturing

Discrete about challenges

“Most IT vendors are extremely proficient at implementing solutions that solve a specific problem: an ERP, a web portal, mixed reality…” says Piet, who is heading delaware’s vertical program for discrete manufacturing. “Via applied innovation, we want to demonstrate how separate tools can work together end-to-end to solve industry-specific challenges – independent from any single technology vendor.”

In discrete manufacturing, for example, these challenges include:

• Hyper personalization: offering customers personalized and relevant products and services.
• Networked industry: breaking down data silos, turning data into actionable insights, and sharing these insights internally and with supply chain partners  (with data monetization as the ultimate goal). 
• One step less, one step ahead: becoming the easiest company to work with and the most attractive employer.
• Sustainability: reducing the company’s ecological footprint, increasing resource and information efficiency, and introducing ‘carbon accounting’.

End-to-end integration via a digital twin

To see how these challenges would map onto specific Industry 4.0 solutions, Piet and Rens created a model that shows how various Industry 4.0 use cases create value together when they are combined and integrated end-to-end – from suppliers and operations to customers.  

According to Rens, the glue that ties all these solutions together is a ‘digital twin’: “A digital twin is a virtual or digital replica of real-world, physical entities, e.g., devices, engines, machines, processes, or even entire businesses. It combines data from different sources, like your shop floor, ERP, CRM, MES, etc. In the end, it helps people make more educated, model-driven decisions.”

Roughly speaking, digital twins serve multiple purposes:  

• Digital representation: what are the current parameters of a specific machine? What data can we gather from a device or an engine? And how does this data relate to other production lines or business areas within my company? Are there any connections we didn’t know about?
• Historical analysis: looking back on a specific timeframe to know what the exact conditions were at the time. By combining various data sources, a digital twin can help you understand why a breakdown occurred. These insights enable further optimization - e.g. in the form of predictive maintenance - which ultimately results in less downtime and improved performance. 
• Simulations: determining the impact of specific changes to your production environment, based on historical data by simulating those changes. Find and understand potential points of failure and act upon those. Or think bigger: before building a new plant or releasing a new product, you could simulate them via their digital twin. In this way, you can check the feasibility and optimize where needed.
• Training: a digital twin of our shop floor or plant allows you to create a virtual environment (think ‘metaverse’) to simulate realistic training situations for your employees. These environments are risk-free, so your employees can learn or see dangerous procedures before doing them. Studies have shown that this immersive way of learning results in improved learning performance.

Collecting data on all levels

Apart from building a digital twin of your factory, it’s also possible to build one of your products. “Data from your digital twin allows you to obtain various points of view on a single product and provide relevant insights at each step in the production process,” says Rens. “This can include, for example, the amount of energy it takes to produce a specific item. That information can then be sent to sustainability solutions that can calculate the exact CO₂ footprint per product.”

So where does the digital twin collect data, exactly? “It all starts with order intake,” says Piet. “Often, this happens via a B2B portal with a configurator. Here, we’re already addressing a key challenge in discrete manufacturing: hyper personalization. Next up is sales order automation. Without any human intervention, orders are scheduled for production and live updates are broadcast. Then comes the actual production, where data on energy expenditure can be gathered and the product passport is generated. The final step is shipping. Here, the product’s relevant data is included on the packing slip.” 

Making it tangible

But what does all this collecting of data and gathering of insights mean on the factory floor? To make the real-world implications of an integrated factory of the future more tangible, Rens and Piet built their own miniature smart factory. 

“People often ask us to show them how a smart factory would actually work, in practice,” adds Rens, who is heading delaware’s key program for supply chain of the future. “Like: ‘what would it mean, exactly, when we’re using data to make automated decisions or perform real-time optimizations?’ The model makes the implications of that immediately clear.” 

Powered by people

“Because we’re starting from challenges and trends in every project, and not from standalone technologies, we ensure that the business is in the driver’s seat on the road to industry 4.0,” says Rens. “In our miniature factory, we’ve mainly been looking at discrete manufacturing for now, but the same could be done with other industries as well. And that’s exactly what we’re planning to do in the near future.” 

Of course, all of this is not taking into account the role that people play in making Industry 4.0 happen. “As we’ve mentioned before, the successful implementation of Industry 4.0 depends entirely on whether your employees are on board, and whether they trust your strategy and the technology. That’s why it’s important to be as transparent as possible about your strategy and its impact. Being able to demonstrate how things would work via a miniature factory can certainly be an advantage, but don’t forget to focus on training and change management either.”