The beginning of the fourth industrial revolution was announced in 2011 at the Hanover Fair in Germany, and the expectations were high. Digital transformation got onto everybody’s agenda, and the vision of having Digital Twins for the entire enterprise, not just the products, appealed to senior leaders worldwide.
Now, 13 years later, almost two thirds of all enterprises are reporting that they are behind with their digital transformation initiatives, whereas others are already embracing A/I in their day-to-day operations. Of course, there are multiple reasons for these very mixed results, but when you take a closer look, it is all about data integrity and digital maturity, and process efficiency.
In this article we are looking at the challenges of a global manufacturer in the classic “iron and steel” sector. The company grew through acquisitions, and the digital maturity of the roughly 50 manufacturing sites varies significantly. Also, this business is quite traditional, and still has to support and even build products that were designed decades ago, and that means tons of convoluted data, such as different BOM structures, missing 3D models, and neglected tracking of engineering changes. It is a bit like a classroom full of children with different abilities, but in the end they all need to succeed, and you must come up with a plan to make that happen.
When we started our digital journey, we decided that we needed to establish a solid digital thread first before we could even think about a Digital Twin. In our opinion a digital thread is just a more modern version of the somewhat dust collecting concepts of a PLM. For us the two terms are interchangeable.
What does PLM mean for us? It is a seamless propagation of data through various systems, even outside of the classical PLM, but with one single source of truth. Since the stakeholders during the life cycle of a product need to see different views of it, there is not just one BOM, there are different BOM’s at various stages of the product’s life cycle, like As-released, As-implemented, As-planned, As-Built, As-delivered, As-Maintained, As-Serviced etc. To maintain a perfect digital thread all these BOM’s should have a single source of truth and seamless connection between each of those. Data integrity of each of those BOM’s is the fundamental factor for the success of digital thread. Even though As-Released, As-Implemented, As Planned and As-built BOM’s are in factory control, most companies are still struggling to connect these BOM’s and establish a single source of truth, and that is because of their legacy practices and processes that they are not willing to give up when new systems are being implemented. It is like bringing home a Ferrari and install a Fiat 500 engine and expect it to still perform like a Ferrari. The challenge is that all the legacy processes must be revamped to support the Ferrari.
With the tools currently available in market, the initial creation of Digital Twins is not the hardest part. That initial creation of Digital Twins is really only half the battle, but keeping them up to date and relevant for the Life Cycle of the product is the most challenging part, and most companies fail at this, unless there is a strong and disciplined ECN change process in place. Only then can the right data at the right time be propagated throughout the Digital Thread. If such a data integrity is not maintained, then the true Digital Twin will remain a distant dream. Therefore, striving for 100% data integrity should be paramount for companies to achieve a successful digital transformation.
Now that we have addressed the data integrity, let us look at the digital maturity of products and manufacturing sites in more detail. To create a reliable Digital Twin of the factory all products that are being run at that location need to be digitized, and that means fixing the past while improving the present. If one is not willing to spend a significant effort on e.g., creating 3D models of an older design one must accept that the digital transformation will only be fully completed when all that legacy products are being phased out.
And one can also not lose sight of the infrastructure on the shop floor. A Digital Twin has the purpose of making a factory more efficient, but also trigger corrective actions at a very early stage. This is only possible if the Digital Twin is in constant communication with its real-life sibling, the shop floor. And this means that the volume of shop floor data puts a high strain on the IT infostructure, an effort that should not be under-estimated. With the ever-growing risk of cyber-attacks, the entire eco systems needs to be constantly monitored and its protection requires full attention.
Companies also tend to neglect the physiological effect on employees, customers, and other stakeholders. It is not just that shop floor workers must get used to new Human-Machine interfaces, the manufacturing engineers in the back office also must accept working in new ways and in new systems. There is always the risk of encountering so called digital fatigue, which is the result of clunky and slow IT systems. If an employee is required to frequently work in different systems and/or a task requires more than 5 steps, or takes longer than 70 seconds, then the users will resent the system and find workarounds*. The risk of digital fatigue can be mitigated by choosing the systems wisely and by automating as many routine tasks as possible. But again, such automation is only possible if it is based on solidly defined processes and impeccable data.
Automation applied to an inefficient operation will magnify the inefficiency. (Bill Gates)
* Source: Wie die digitale Ermüdung die Digitalisierung gefährden kann (walkme.com)
Susanne Lauda is the Director of Global Advanced Manufacturing Technology at AGCO Corporation
Anvesh Kulkarni is the Global Business Process Lead for Manufacturing PLM at AGCO Corporation
