The “Industrial Internet of Things” (IIoT) is a topic permeating all aspects in the production of tomorrow. Yet it often fails to extend beyond the purely theoretical. This is where we ask a very specific and practical question: What can and should you do now?
As you’re probably well aware – forecasts about the forthcoming digital revolution abound like sand on the beach. “Everything that can be digitalised will be digitalised”, stated the Federal Chancellor some time ago. Industrial production is becoming increasingly shaped by this development. However, it sounds like a vague vision when you talk of more flexible production systems and falling unit costs as promised with Industry 4.0 and the IIoT. And manufacturing companies from Germany are indeed lagging behind according to the PwC study “Global Digital Operations”. According to this, only seven percent have a high “digital ecosystem maturity” – in the Asia-Pacific economic area this is 15 percent however. My impression from practical experience: There are many IIoT platforms in Germany with no shortage of good ideas for production. Unfortunately, the topic does not enjoy any special significance in companies. “What specific actions can and do I have to undertake now”, is a question posed by many medium-sized companies from manufacturing and mechanical engineering. I would like to proffer a few answers to this. The first section will deal with the establishment of a promising business model. The second section will focus on the technical approaches and platform required. This division enables a three-step course of action: Only once the business model has been determined should you create the associated IoT application and then choose a technical platform with which to optimally realise the application.
The first challenge for companies in this topic field is therefore the development of an IIoT-based business model. At this point we should emphasise that we are not inevitably dealing with something revolutionary. Rather two fundamental approaches are involved:
All cases call for a viable business model that can provide the application with a promising framework. Its development is not a task that companies can achieve as isolated lone wolves. Quite the contrary: Digitization is teamwork – extending beyond company limits! Experts talk of an extensive co-innovation here, in which even a large conglomeration of companies might collaborate. It is also important to follow scientific methods when developing the business model. They have two common denominators: Teamwork and openness. These words call for really radical thinking. It is a matter of absolutely open exchanges with one another, while continually questioning your own work and learning from mistakes. In this context, wisdom is not written in stone – rather there is a “validate learning” whose insights are assessed and if need be corrected time and time again.
lDigitization is teamwork – extending beyond company limits! Experts talk of an extensive co-innovation here, in which even a large conglomeration of companies might collaborate.r
Many methods for developing business models originate in the start-up field and are not least the result of inadequate resources. Nevertheless, established companies can adapt them for themselves, as these methods are extremely helpful depending on structure and target objective. To begin with, the “business model canvas” would be worth mentioning here – in other words a “business model on a blank canvas”. Put simply, this is a matter of presenting your own business idea on just one large page in a very early phase. For instance, have you come up with the idea of customising your machines and systems using an automated “Configure-2-Order”? Record the central steps from the configuration and parts manufacture required for this through to delivery of the machine! The result is the focus of a completely open group discussion, its fluid nature allowing it to be shaped and reshaped.
The so-called “Minimal Viable Product” (MVP) strikes a similar vein – only this does not involve openness in the business plan but instead in the product development phase. The new machine or software product is developed at less expense and then offered as a prototype. We could cite a B2B sales platform here as an example, which initially has just a few functions for test purposes or is only available to a limited user group. The number of users is then counted – a field test with reliable statements regarding future potential. For instance , the MVP approach was recently chosen by a Hamburg start-up, which offers online booking systems for logistics companies. They utilised an inexpensive beta version of their platform to find out whether logistics providers are actually interested in offering services online.
The interface between man and machine forms a critical pint in the IIoT. It is precisely the “pain points” that determine how effective the overall solution will be – in other words how the user’s problems can be remedied in the machine? This issue is at the centre of “design thinking”, in which the needs of the user are always treated equally with technical aspects. To this end, people from different disciplines work together in a combination involving understanding, observing, refining and continuous learning. Step by step, they come closer to designing a product that not only functions technically but is also user-friendly to the operator. This is the cornerstone to any market success, because the “usability” of the machines affects the productivity of the entire site. Ultimately “design thinking“ is thus relevant for the individual user as well the person responsible in the company alike.
What is also relevant: The development of digital products needs to progress at a rapid pace, as technological development occurs in ever shorter cycles in today’s age. The processes involved in software development cannot be compared with the time periods required by a mechanical engineer to develop a new generation of machines. These are only a matter of a few weeks! “Pivoting” is an interesting model in this context. This states that a product is not immediately and inevitably finished if it fails to make inroads in the market. On the contrary: Failure is the starting signal for something new. You need to make an about-turn and possibly alter your business plan. The history of major companies is replete with speculative pivoting examples. Thus Steve Jobs started out by just selling computer accessories before he radically transformed his business model to found his own brand.
Ultimately, the central factor in success – “development speed” – is influenced by many things – yet people and ways of working play a role in all cases. Interestingly, diverse examples from high-tech bastions like Shenzhen or Silicon Valley reveal how helpful it is if you divide a product development into stages and then have the responsible team work as long as possible on the project in so-called “sprints”. Afterwards everyone takes a longish break and attends to preliminary work before tackling the next sprint. It is interesting to see that even traditional groups of companies are adopting this approach, like a German electrical appliance manufacturer for instance: In the past it took about one and a half years to reach market maturity here. Now agile teams provide prototypes after just a few months.
Of course “Business Model Canvas”, “Minimal Viable Product” or “Design Thinking” cannot be applied like templates in an established company. Ultimately, it is indispensable to customise your own expertise and the application fields individually so as to develop a viable framework of action building on this. Another rule of thumb: Methods are one thing, technological expertise is an entirely different matter. The second factor cannot be outweighed by anything. You need cutting edge research, high-tech knowledge and experience plus an entrepreneurial spirit coupled with a willingness to take risks. This can result in functioning business models that rival the unbelievable success stories behind start-ups in China, Israel or the USA – and precisely this is the benchmark.
Once the business model is in place, it’s time to look at the technical challenges associated with establishing IIoT solutions – a broad field with a central task: creating a comprehensive link between machines, systems and devices as well as the cloud, while embedding everything in a stable, clear and rapid architecture at the same time. The second factors plays a crucial role in particular, because a large number of central “players” are involved. The data transport, its protocolling and transformation as well as ultimately its processing into an IIoT application call for very clear structures.
What needs to be said straight: The practical problems arising from a large number of networked machines are underestimated. Firstly, no two factories are alike. Is the production facility fully automated or are there predominantly semi-automated processes? Do many of the machines used have similar processes and interfaces or is each of them an individually designed unique product? These and similar questions influence how Industry 4.0 networks are established. Plus it is often the case that the machines are up to 30 years old and not available to any OPC-UA server (Open Platform Communications Unified Architecture). The jungle of words conceals a collection of standards for communication, data exchange, data transport as well as the structure of data, interfaces and security mechanisms. The original OPC specification appeared as early as 1996, although it only took a crucial step forwards in 2006 with publication of the first version of the OPC UA – the reason: The machine data are no longer just transported but also described in machine-readable form. In this way, OPC UA enables communication between the products of various manufacturers. This is the cornerstone for a comprehensively controlled smart factory.
But is it actually possible to establish IIoT approaches in older production environments against this background? The answer is fortunately “yes”, as there are many machines and systems that use programmable logic control (PLC). With a number of obstacles, their language can be translated into the OPC-UA standards – and here too experts with detailed practical knowledge are needed. However, this is only the case when you have access to the original PLC programs. But there are two snags here: You need to work together with the PLC manufacturer – if the latter still exists – and the entire process entails costs. Another option is to integrate new sensors in older machines for use so as to equip them ideally for the IIoT. In the last few years, a number of manufacturers have developed Industry-4.0-compatible sensor solutions that communicate with the cloud via OPC UA. As a rule, every path leads to OPC UA as a “world language” for machines. Yet we are only at the start of a long road ahead of us. Once adopted, this internationally standardised information architecture will enable users to “understand” the information model for a machine type comprehensively in future – but it will take years until all relevant accompanying specifications are formulated.
A keyword should not be absent when establishing IIoT applications: Big data. When operating an individual (!) machine tool, a large volume of sensor, operating and production data is accrued. But we’re not talking about one machine, measuring instrument or production system here. The overarching vision is to network a complete production site or even multiple production locations with one another. After all, these diverse “players” not only exchange a wide variety of data via the Internet – instead it involves diverse internal systems that have to interpret and analyse huge volumes of information in real time. The transported data volume therefore increases exponentially. It is obvious that high-performance hardware and software is indispensable if a rapid data flow is to be guaranteed even at load peaks – and precisely this is an important investment field you need to keep in mind right from the start.
We should begin here with a clear statement: Cloud solutions are the future! Why should it be no different when the heart of the matter involves networking the production processes of a globalised world together? Cloud solutions also offer an ideal technical basis for many companies because they can be scaled with relatively little expense if companies grow and alter their locations, processes and production systems over the course of time. Cloud solutions including their usage models offer tremendous freedom here. Plus the financial entry hurdles are low. A typical criticism of the cloud – above all in Germany – you’re better off writing hardly anything down: What about security? A justified objection, but one that can be countered just as quickly. On the one hand there are security-certified providers with local data centres. On the other hand, the world “cloud” can be understood flexibly here, as the solution can involve an on-premises and off-premises solution. This guarantees all-round security for particularly sensitive data. This leaves us with the question of choosing the right provider. A generalised recommendation is certainly impossible. Major players like Microsoft, SAP, Amazon Web Services or Google offer their own solutions – according to market studies more than 400 platforms are available in total. Making the right choice concerns the project, its objectives and the resultant technical specifications. Independent know-how is yet again indispensable at this point. The entrusted experts should contribute extensive experience from other IIoT projects to offer advice building on this.
The traditional business models of many manufacturing companies and mechanical engineers will come under pressure in the next few years – for instance because their “analogue” products are at the end of an innovation cycle and new competitors with digital innovations are penetrating the market. Whoever wishes to see how development could take shape need only look at the already radically transformed business models of banks, booksellers or insurance companies to gain an impression. Your sector has not been affected so far? Things are changing. On the other hand, there is no reason to bury our head in the sand, as many established companies have tremendous expertise in their application fields – only so far they are making precious little effort to develop and advance this knowledge for the coming IIoT age. “The only way to win is to learn faster than anyone else”, as the ‘Lean Start-up’ pioneer Eric Ries once stated. We quickly need to get to grips with learning better.