Surveys and market analyses currently assume that around 40% of medium-sized
medium-sized companies have implemented or are in the process of implementing IoT projects. This contrasts with analyses by McKinsey and Forbes, which assume that over 50% of these projects fail or do not lead to the desired technical or economic success. Why is it that the majority of medium-sized companies have still not jumped on the bandwagon? And why do so many of these projects fail?

The reason why in-house implementation often fails is due to incomplete requirements, poor use cases or overly optimistic and naively planned business models. Projects are discontinued after a while or solutions do not make it beyond the boundaries of a specialist department. The reasons for this are often: too specific, too inflexible, too expensive and/or too complicated. As many SMEs do not have the necessary resources or the required expertise, external providers and service providers are often brought in or commissioned with the implementation after the first unsuccessful attempts.

Is this a suitable scenario?

In order to bring projects and digitalization strategies to a positive conclusion for providers and users alike, clients and contractors should be able to answer fundamental questions. Due to their complexity, (I)IoT solutions usually require a comparatively high level of explanation.
Especially in the absence of experience, a suitable scenario with manageable complexity should give the customer the opportunity to easily get to grips with the topics relevant to them and build up their own experience and expertise.

What measurable value should the solution offer?

Projects often fail due to flawed criteria, unsustainable business models or simply a missing problem, because not everything that is technically possible necessarily makes sense or appears relevant and beneficial to an end customer. Regardless of whether a project aims to digitally enhance existing solutions, new products or services - without a quantifiable benefit, any solution, no matter how well implemented, is reduced to a gimmick. Concrete specifications not only define a framework for the effort and costs of a solution or additional investments for the required technology and monthly operating costs, but also define clear, tangible and comprehensible limits for the technical and economic success of a solution.

Which usage scenarios should be realized?

André Sachs is a developer at Gett Gerätetechnik.

© Gett appliance technology

As diverse as the usage scenarios are in the age of Industry 4.0, so too are the possible solutions and their respective strengths and weaknesses. The requirements and objectives of after-sales services, condition monitoring, predictive maintenance or X-as-a-Service solutions, for example, could hardly be more diverse. Success-critical factors such as availability, data volume and quality, functions, flexibility or operating costs must be weighted and prioritized accordingly depending on the application. In this way, expensive over-engineered solutions can be avoided, as can low-cost solutions that do not meet technical requirements such as scalability or flexibility.

Which data should be used and how?

Even if the transfer and storage of data is becoming increasingly cheaper from a historical perspective, a suitable data strategy is essential for the long-term success of a solution. Depending on the usage scenario and possible future use cases, it is important to plan the volume and quality of data in order to reduce the effort required for collection, transfer and processing to the necessary minimum.
Especially with regard to machine learning applications, too much unfiltered data as well as incomplete data can lead to problems that can affect the training of algorithms and later applications.
At this point, it can be helpful to think about the value of the collected data and its further use. How can the value or information content of the collected data be increased? What additional information can be derived from it? This can open up additional opportunities for new applications.

What costs need to be kept in mind?

In addition to the costs for implementation and integration, the follow-up costs for operation and support can also be decisive for success. Depending on the use case, technologies, platforms and infrastructure used, the return on investment can be a long way off. The profitability of a solution or an entire product quickly depends on transfer costs for data, fees for platforms, licenses for services, interfaces and software as well as support costs.
Particularly with regard to costs for platforms, support and services, supposedly inexpensive entry-level offers or all-in-one solutions can develop into significant cost drivers as the amount of data or number of users increases. A transparent, manageable and predictable cost structure is therefore essential.

What resources are needed?

In order to be able to think long-term and develop successful strategies for projects and, not least, minimize possible risks, it is necessary to keep an eye on all the resources that are already available and any additional resources that are required. In addition to all kinds of technical resources such as sensors, infrastructure, gateways and interfaces, human resources must also be included in the planning.
The implementation of Industry 4.0 projects is an interdisciplinary challenge that often requires specialists, knowledge and experience, especially for industry-specific solutions, which cannot be easily outsourced or bought in and out.
If the necessary expert or detailed knowledge is suddenly no longer available due to inadequate knowledge management, or if a host of IT experts are required for operation and support in the long term, even solid projects can go awry.

What are the long-term goals?

Regardless of whether new functions, products, services or complete ecosystems are to be created, the course for a high-performance and affordable solution must be set sooner or later depending on the application scenario. Future requirements such as scalability, data, user and device management are not insignificant for solid project planning and costing. Clear strategies and priorities must be defined, particularly with regard to current and potential costs and requirements.
Forecasts, empirical values and rough calculations can already provide good indicators as to which strategy is best suited to achieving the set goals.

For example, a redesign or an expansion with additional modules for new functions can be significantly cheaper from a certain number of units than relying on an expensive all-in-one solution from the outset. Or the migration from platform provider A to platform provider B is only worthwhile from x-thousands of users.
Developments in recent years have put us in a position to choose from an
to implement suitable, affordable and standardized solutions for almost all areas of application from an almost endless portfolio of hardware and software. Open standards, interfaces and ecosystems are the driving force behind advancing digitalization in many areas. Even if one trend follows another in the area of I4.0 and (I)IoT, it is always important to develop the right approach and the right strategy individually.