Automation in manufacturing: how to design for efficiency and ROI

Technology has opened up new horizons for us: so much more is possible with automation now compared to 20 years ago.

Skilled workforce is the main challenge industrial manufacturers are facing today, and it's also what's driving automation. We need skilled people that can program automated solutions. Programmers need to be able to program AI and understand various technologies, product development and production. So it’s safe to say manufacturing is getting more complex.

Another challenge is the rapid pace at which technology is evolving. Any sort of direction that we choose to take today might not be valid in ten years, so flexibility and perspective are important.

Most manufacturing companies are still heavily siloed. This leads to production engineers solely focusing on production and R&D focusing on designing the best product for the customer. However, it’s good to keep in mind that the best product for the customer is actually the one that is the easiest product to assemble.

So, to keep the cost of automation down, it’s crucial for process engineers and R&D to start working together a lot sooner. To facilitate this, company culture also needs to shift towards thinking of not just design but making sure the design is practical to assemble as well.

Another issue that manufacturers are facing is that many assembly lines don’t have enough space for automated stations. The traditional way of assembling a car, for example, is that you first start with the auto body and then move inside the car, which is not practical for automation. This is where unboxed manufacturing comes in. New technologies enable us to manufacture in a new way, like with gigacasting where you can produce big aluminum parts and easily sub-assemble them, avoiding obstacles like going inside the car.

The bottom line is that eventually linear production will reach a point where it cannot be more efficient than it already is. Manufacturers who want to keep their competitive advantage will therefore have to rethink their assembly processes. And this doesn’t just concern the manufacturers: suppliers will also need to make sure that they can provide the right components and the right materials. Automation is challenging the whole system and requiring all parts of the chain to work together.

Despite the challenges, there is a lot of money to be saved by getting automation right. Here, simplicity is key to reaching your target ROI.

Often manufacturers think they need to go all in, automating either 100% or nothing. However, for companies that are just starting their automation journey, it can be easier and still worthwhile to get started by automating just some parts of the bigger picture.

Like so often in design, simplicity always looks simpler when you look at the end result – but the magic lies in figuring out how to achieve it. Designers and engineers should be working together to reach a simple and lean process because that will also be a sustainable one, and the one that saves the most money.

One good example is the packaging industry. In order to achieve high efficiency and the high level of automation typical in the industry, it’s often the case that the packaging machine builder and packaging designer are the same. Either the packaging machine company designs the packaging to suit the machine, or the packaging designer also produces the machines for their packaging. This demonstrates that the only way to achieve such a high automation level is by production and design to be operating in a very close loop. The machine and product need to meld together and be one.

At Atlas Copco, our philosophy for automation is called “nail it and scale it”. Around three years ago we started to really put extra intention into how the tightening process can be automated very efficiently. To achieve this, we created our own team with a not so typical R&D setup. The key is to be close to the market and close to the customer, to really understand their needs and start from there with digital simulations. Digital twins help with very early concepts and figuring out where the optimization already can be done together with the customers and all the stakeholders. We can then identify the critical components and move on to physical prototyping with confidence.

This is the always starting point for us before we even start with development. It should always have the aim of being scalable, so that customers get a really proven solution in the end.

Digital twins are a key resource for the design process. With a good digital twin, you can really understand what's going on. You can switch things around, test out things in the virtual setting and understand the actual repercussions, communicating them easily to all relevant stakeholders.