CREATE FLEXIBLE SMART MANUFACTURING OPERATIONS
For decades, A&D manufacturing processes have been geared towards the production of single or small groups of products from centralized facilities, using fixed assembly lines and largely manual and labor-intensive processes. Their specialization has made it hard for such facilities to quickly adapt their product offering, or quickly scale in response to demand. One example of the limitations of this approach was noticed in 2020, when the Trump administration invoked the Defense Production Act. Many A&D OEMs were now required to produce things (like ventilators) to aid in the American public health response to COVID-19, but not all of these OEMs were able to make the transition smoothly.Smart manufacturing exploits new technologies – like AI, robotics, additive manufacturing, and the Industrial Internet of Things (IIoT) - to produce a greater diversity of products, more quickly and more efficiently. In a world where demands change rapidly and multi-year lead times are not good enough, A&D will need to embrace these technologies to become more flexible and adaptive. To do so, they should consider the following.
The share of organizations with a long-term roadmap for digital twins increased from 57% eighteen monthsago to 73% now
We can think of the above as a move from a ‘project organisation’ to ‘platform organisation’. Rather than setting up a facility to deliver a product line, we are setting up a platform that can be easily adapted, through changes to software and hardware, to deliver a wide range of products - bridging the physical and the virtual world and mutualizing resources, whilst driving design and operational efficiencies.Such a platform can also connect up to other systems, such as engineering and supply chains, allowing engineers to run models of new designs on simulations of your real manufacturing environment and, once approved, quickly translate new designs into manufacturing setups and instructions, without having to move through silos.Getting this right will need work to create consistent data and models that work across the organisations – between design and manufacturing and between plants – which will require data standardization, ontology definition and so on.The value of maintaining a ‘digital thread’ of this platform data and ‘digital twins’ of its various components and products is well understood across OEMs, suppliers, operators and space companies. In fact, our latest Capgemini Research Institute report, Mirroring reality: Digital twins in aerospace and defense, reveals the share of organizations with a long-term roadmap for digital twins increased from 57% eighteen months ago to 73% now.
Such transformative endeavors are difficult and are unlikely to succeed without an informed plan of action. This doesn’t necessarily mean starting afresh with a new greenfield site. Think of it instead as an incremental process of ‘smartening’ your existing capabilities, allowing you to do more with your existing infrastructure.
Clearly outline the objectives you want to achieve with smart manufacturing. Analyze the existing challenges and pain points in manufacturing and the supply chain that you wish to address. What are the risks (eg. cyber, political, skills shortage, legislative changes and sustainability) and how will you mitigate them? What will constitute a positive ROI? Audit your infrastructure - will you need to modify existing systems, or do you require new ones?To become more responsive, new setups and facilities are inevitable. But misjudgments in their design can cause even more delays. This is why it’s important to simulate new approaches first.
For example, dynamic flow simulation can be used to model how a factory setup will work, visualizing how parts will move through a production line, and allowing you to optimize setups. We worked with a company to model a new production facility, factory resources, machinery, and people flows to get a changeover cycle time between takt windows from 30 minutes down to the target of 15, by changing the rules and equipment used.Iterate all of this in fast but manageable steps with stage gates to learn and adapt. If we’ve learned anything from the technology sector, it’s to make ‘small bets’, and quickly learn from them.
To this end, you could try ‘dipping your feet’ in a smart manufacturing proof of concept. For example, can you test a ‘dark shift’, where your factory runs overnight using automation, entirely without people present? Or can you trial a limited introduction on a line? This way, inevitable mistakes can be made (and corresponding lessons learned) on a small scale - before being applied to the business as a whole.
This may sound like a recommendation from ten years ago, but many shop floors still have lots of legacy manual technology. To increase production outputs and become more flexible, they must first automate as much as possible.Evaluate the solution space. Which legacy manual systems will be easiest to automate, through replacement or the addition of connectivity? Take the example of logistics. In a relatively simple intervention, Automated Guided Vehicles (AGVs) or more sophisticated Autonomous Mobile Robots (AMRs) could be used to automate the movement of goods and materials between lines or cells.
In this context, ‘flexibility’ can be thought of as the ability to move from one product to another very quickly. A&D production lines are typically built around producing a single product. Yet, as demands change, factories can benefit from more flexible setups – for example, a sudden rush of orders for small planes and a drop off for large ones may leave some factories overburdened whilst others twiddle their thumbs.Being able to efficiently assemble a diversity of products requires manufacturing flexibility. This can be aided by late diversification. This describes a production method in which products are customized to match the customer configuration at the latest point in the industrial process - which can simplify production, decreasing the number of specialized parts, and the need for sub-assemblies.Smarter setups can help. For example, the use of AI tools that can plan, using analytics, to avoid blockages of production lines. Flexible production techniques such as robotics and 3D printing can be adjusted on the fly with software. A digital twin of a production setup can then be mapped from one factory to the next, to change what is being made.Doing this means investing in production systems capable of being adaptive, connecting them up, ‘softwarizing’ them (ie. creating digital versions on the network that can be controlled through digital interfaces) and designing a network that connects them all intuitively, so that whole new setup can be easily programmed. Be sure to consider system integration when investing in new systems for your lines - are there open standards that will ensure ‘out of the box’ compatibility as you introduce new capabilities - or will further work be required? Assess whether you have in house skills to do this, and, if not, identify new employees or partners (like a system integrator) who can deliver it.
Where possible, design your products to be produced from a diverse range of parts, materials, and locations.Modularity and interoperability aren’t just buzzwords, they’re ways to build in flexibility. By designing products that can be initially produced as modules, then assembled in the final stages, you increase the work that can be done outside of final assembly lines, building much more flexibility into your process. A number of major aerospace companies have already embraced this approach, redesigning their industrial processes around the major component assembly (MCA) concept - in which major components, like the fuselage and wings, come from suppliers mostly preassembled - making the manufacturer’s job easier and faster when completing the final product. Thought should also be given to the ‘design to manufacture’ approach - a concept that breaks the siloes between engineering and manufacturing, merging process and product design. The benefit of this approach is that the design process doesn’t just optimize the platform, but also the production of the platform - ensuring that manufacturing a new system is as efficient and requires as little retooling as possible. Put another way, avoiding the scenario in which a company designs a sophisticated product that is too sophisticated to produce effectively.
Sharing data with other business is not a default behaviour for A&D actors. There are very real concerns about customer privacy and IP. However, if done correctly, companies can share and collaborate to help build a more resilient A&D manufacturing and supply chain environment.Where possible, embrace cloud-based digital engineering platforms for better communication of needs, specifications and relevant data (eg. forecasts or demand information) to suppliers. There are some existing examples that may be worth looking at - for example, the UK Ministry of Defence’s Engineering Delivery Partner – a route for procuring engineering services for various MOD departments and agencies. Use the data model you’re building to create a situational awareness picture of your supply chains. Develop mitigation strategies to address these risks and establish contingency plans - for example, what will you do when (…not if) the next pandemic hits, demand rapidly declines, and suppliers fold? Or if conflict engulfs an area you rely upon for vital services or materials? Consider also the growing importance of cybersecurity, as supply chains grow increasingly digitized. Again, we reiterate the importance of a crisis management strategy - not just for your supply chain, but also in your manufacturing.
