Q&A | Building a smart factory from the ground up, with Lilium’s...
Lilium Jet

Q&A | Building a smart factory from the ground up, with Lilium’s Thilo Braun

Lilium is developing the world’s first electric vertical take-off and landing jet. It is part of a company vision that aims to bring affordable air travel to everyone, through an aircraft that they claim will take passengers to any destination five times faster than a conventional taxi.

With over $100 million in funding raised so far, including the likes of Chinese tech giant Tencent, Lilium has quickly gained traction as one to watch in the transportation space.

We spoke to the Head of Industrial & Production Planning at Lilium, Thilo Braun about how they are going about building the scalable, smart manufacturing capabilities they hope will enable The Lilium Jet to change travel forever.

Internet of Business: Could you start by telling me about your background in smart manufacturing and your role at Lilium?

Thilo Braun: “I come from an aeronautical engineering background. My first involvement with manufacturing was at Lufthansa Technik, which was completely paper-based, and I worked to introduce the first data acquisition system.

“Going through this process made me realise just how people-centric smart-manufacturing is. Much of my time was spent getting user-acceptance, making sure it’s actually getting utilised at the end, and involving the users from the very beginning of the design.

Thilo Braun - Lilium
Thilo Braun, Head of Industrial & Production Planning at Lilium

“Fast forward to Lilium and I’m heading the industrial department. The vision for us is to build a truly digitally-native production system. On the one hand, we’re putting the users, our employees, at the centre and moving away from any legacy systems. And on the other, we’re thinking about how we can implement digital systems in a way that will help the people working in the factory and improve the efficiency of the system.

“That can mean management tasks – things like predictive maintenance – but also just generally improving transparency at that end and making processes faster.

We’re considering several options. Something that hasn’t received a lot of attention yet, particularly in the aviation industry, that I think could make a huge difference, is adaptive work instructions – looking at non-conformance processing and how that can be done adaptively.

What are the key elements in your approach to building a smart factory?

“From the very beginning digital twin technology has been a vital aspect, not only for the final product but also the production system – both in terms of the design and, later, the operation.

“Digital twin is a very wide term, so to break it down a bit, we started by creating a sort of visual digital twin of the factory – something that isn’t really a dynamic representational model but merely a representation in 3D. This was incredibly useful for getting user feedback.

“It enabled us to have investors and manufacturing engineers walk through the factory and provide their feedback. Crucially, it also allowed the production staff to see their workstations and help us improve the factory design from there.

Beyond that, we’re looking into doing simulations of how a supply chain responds to failures and unforeseen events.

How does this feed into what you’re trying to achieve with the Lilium Jet?

“The vision behind The Lilium Jet is to offer air mobility for anyone, anywhere, and at any time. Translating that into the production means going far beyond the normal production volumes for commercial jets. In the end, using a digital approach to factory design, as well as operation, gives us the potential to achieve a much faster ramp-up curve and a far higher likelihood of achieving stable production at high volumes early on than we would have without.

Lilium Jet prototype
The company has flown its two-seater prototype of the jet in the skies above Munich

What are the advantages of being able to design a factory for Industry 4.0 from the ground up, rather than having to retrofit?

“The largest advantage we have is we can design the entire system without any constraints. It’s like building a factory. When you build a greenfield factory, you don’t have any of the physical restrictions that you do when building a brownfield one. The material flow can be much more efficient, materials more readily available, the paths for labour and other logistical factors better arranged, and so on.

“Of course, you can get close to these kinds of efficiency gains with a brownfield facility but there’s always going to be constraints. Looking at the digital system, it’s very similar. If you have a brownfield, you’re looking at having data storage and data flow that is already established in a certain way. Yes, they can be changed, but that’s going to cost a lot.

“At the same time, you have systems and processes already in place that you have to work around or replace. But you’re always going to be looking at building interfaces, which then create hard constraints and start to limit the flexibility of what you’re doing, or significantly increase the work required.

“So for us, having this clean-sheet-thinking approach actually enables us to, firstly, think about how we can design the system to be as efficient as possible from the ground up and, secondly, knowing there will be more we want to do in the future, set it up for flexible expansion.”

Has anything surprised you along the way? Or is there anything you would have done differently?

“One thing that surprised me is the amount of technology out there on the consumer market that, particularly on the software side, is significantly more progressive than what you find in an industrial context. Generally, in the industrial context the user is a bit of an afterthought, so software is designed for functionality rather than user-friendliness.

“This definitely makes sense in many ways, but the user has often been too neglected to make systems widely accepted. At the end of the day we want to make work a great place to be, a place our employees want to be. It can’t help when you have IT systems that are more of a hinderance than anything else.”

What advice would you give to other manufacturers looking to design a smart factory?

“I think the most important factor is to think about processes and functionality first and avoid getting wrapped up in established buzzwords and acronyms. Instead, come from the perspective of what happens in the facility.

It’s also important to value your time and think about the actual work being done in your facilities before diving into anything technical. It’s easy to jump the gun.

What technologies do you see gaining more traction with smart manufacturers over the next few years?

“I think using VR for training purposes is currently under-utilised. There are a few cases of it having been implemented effectively – for engineering design, in production and design facilities, it’s being used quite well already, but on the training side I haven’t really seen a good execution of VR.

“If you look at the status quo, the average production site is really not very digital. Companies are only just starting to use manufacturing execution systems (MES) systems, for example.

“The data quality in production facilities is going to increase significantly but so is the amount of data. One of the biggest problems in actually dealing with these quantities, the Big Data, is that many manufacturers can’t see the wood for the trees.”

Is there anything that you’re particularly excited about at Lilium?

“I think it’s incredible to have the opportunity to start from scratch and see where the things we already have today can take us, but opportunities remain relatively untapped. We’ve come across a few, such as adaptive work instructions, that can be developed in the future to make production more efficient, and to better harmonise the people and technology behind it.”

You can hear more from Thilo Braun on digital twin, and designing a smart factory at Internet of Manufacturing 5-7 February 2019, Munich, Germany.

Internet of Manufacturing DE