Intro

We at ENLYZE have been able to digitize hundreds of systems in the last 5 years and provide insights into the manufactured reality of medium-sized manufacturing companies through machine data. In this article, we share our experiences in the field of digitization in the form of 5 typical mistakes that medium-sized businesses should avoid.

Error 1: Low Data Quality Due to Manual Entry

Once digitization is underway, many companies lose sight of the actual goal of digitization: creating a data foundation that enables better and faster decisions. This is intended to lead to increased revenue through enhanced productivity, reduced costs, or new business ideas.

The “Paperless Manufacturing”, for example, is a good first approach, but is almost always implemented incorrectly. Existing processes are digitally copied one-to-one, with information previously noted by workers with pen and paper now recorded digitally in a MES.

It quickly becomes apparent that the captured data does not reflect reality. Downtimes are not recorded accurately, and orders are not booked correctly. Due to manual bookings, mistakes are inevitable. The poor data quality leads to low trust in the dataset, which in turn leads to limited use of the data for decision-making. A vicious circle emerges: Because nobody uses the data, even less emphasis is placed on correct data capture.

Our approach focuses on creating a solid, trustworthy data foundation through automated machine data acquisition as the first step of digitization. A large part of the manual bookings becomes obsolete, revealing the manufactured reality. Data forms the foundation for good digitization, and if this foundation is crumbling, all decisions based on it will also become irrelevant.

Image 1: The vicious circle due to poor data quality

Error 2: Data Is Held in Silo Systems

To work meaningfully with machine data, context information such as order and product information, reasons for downtime, etc., is required. Only when this information is available in combination with process parameters can it be worked with meaningfully and efficiently.

Typically, this data is distributed across several systems. For analysis, these must then be painstakingly gathered from many systems, exported, and then compiled into a uniform dataset.

This manual process makes analyses not only extremely time-consuming but also very error-prone. This high effort ultimately leads to low motivation among employees for independent analyses and consequently to few data-driven decisions.

If instead all data is in one place - in a Single Source of Truth - all necessary data is readily available. Analyses that previously took several hours can now be done in a few minutes. To enable quick and efficient analyses and thus regularly obtain data-driven decisions in manufacturing, your data silos must be broken down.

Image 2: All data should be accessible in a Single Source of Truth

Error 3: Waterfall Model Instead of Iterative Approach

Digitization means something different for everyone in the company, which is why prioritization is often difficult. Frequently, roadmaps and lengthy requirement specifications are created to meet the demands. These then overwhelm when implementation begins, and the hoped-for goals are not achieved. The problem: initial assumptions are made with little insight into the data and thus into the real issues of manufacturing.

At ENLYZE, we therefore start without digitization roadmaps and even eliminate existing requirement specifications at the beginning of our projects. Instead, we plan in smaller steps, guided by data and new insights, allowing us to build the digitization journey much more accurately and based on facts. Instead of introducing a handful of industry-standard systems like ERP, MES, PDM, our clients can selectively identify problems and address them with the necessary steps and systems.

Image 3: An iterative approach leads to faster and better results

Focusing on this iterative development of use cases also quickly reveals another problem with current digitizations. Many providers sell concrete solutions for existing problems, which are often extremely inflexible for adjustments (at least not without significant additional costs).

Error 4: Focusing on Concrete Solutions Instead of a Flexible Infrastructure from the Start

These comprehensive systems from renowned manufacturers are one-size-fits-all solutions in a modular system. They must be specifically adapted to your requirements and processes in lengthy, costly projects, usually accompanied by large consulting firms with exorbitant hourly rates. You shouldn’t even think about post-implementation changes.

A true alternative is the best-of-breed approach. Here, you choose the best solution for a specific problem, such as an ERP system focused on your industry. Our approach goes a step further. Besides initial use cases out-of-the-box, we offer a completely flexible data infrastructure. This allows new use cases to be developed iteratively and the system to be adapted to your individual needs.

For such a system to truly work, data provision must function flawlessly. This is our experience. Particularly challenging in this regard are heterogeneous machine parks, various connectors, the integration of existing data sources (sensors, controls), and the integration into your third-party systems. 

We advise medium-sized companies against building this IoT infrastructure on their own. The establishment of stable data recording and monitoring of the entire infrastructure is grossly underestimated at the beginning. A managed data platform with advanced connectivity takes away the complex tasks for a continuous flow of data and allows you to focus on rapid, genuine added value.

Image 4: Maximum connectivity and integration of all data into third-party systems through ENLYZE

5. Wrong Driver of Digitization Initiatives

Clear responsibilities between the departments involved in the digitization of the shop floor are also important. From our perspective, the production/operations manager is the ideal driver of a shop floor digitization project.

The production manager knows manufacturing best, and thus specific requirements and the greatest levers for increasing productivity are identified. If the benefits are visible to him and the project is driven by him, the system will also be better accepted on the shop floor. Ideally, a platform is even found that allows operations to make adjustments independently, e.g., adding new data points or determining reasons for downtime.

IT should definitely be involved in the process but should not take the leading role. They handle security approvals, discuss adjustments to the network infrastructure, and implement these. The innovation department, if present, takes care of project management in the initial phase by screening potential providers. 

Outro

Have we triggered critical thoughts about digitization in your company? Are you asking yourself whether you are on the right track to make better and faster decisions in the future?

We look forward to discussing concrete solutions and procedures with you. If you want to learn how to understand the manufactured reality of your shop floor in under 4 weeks, please feel free to contact us via email: hello@enlyze.com.