AI beyond the buzzword bingo

How artificial intelligence is sneaking into every day smart manufacturing applications delivering true business value

The discussions on promising applications of artificial intelligence in operations are coined by extremes. On the one hand, discussions are about highly sophisticated optimization applications e.g. around complex demand forecasting and predictive maintenance. Looking at the reality of most manufacturing companies and the availability of trustworthy data, those applications are only of limited relevance. On the other hand, inflationary used marketing slogans are flooding the market proclaiming AI inside almost every piece of software. When having a look behind the scenes, what has been praised as AI from the outside is often not better than excel from the inside.

From a user perspective, both extremes are not helpful as they do not focus on delivering business value in running operations through the weathers of daily business. Therefore, predicting future benefits of AI in smart manufacturing often feels like looking into a glass bowl and doing alchemy magic. Nevertheless, there are good ways to make more robust predictions on how AI will deliver business value to operations in the future. One way to do this is to have a look at domains that are showing a quicker technology uptake and shorter innovation cycles compared to the manufacturing domain.

Business value delivered by AI in software development

Software development has taken up Artificial Intelligence quite quickly. AI in the form of so-called AI co-pilots are common features in programing frameworks supporting developers in fulfilling their daily tasks. This is due to three major benefits provided:

Increased Productivity: AI co-pilots can automate repetitive and time-consuming tasks such as code generation, refactoring, debugging assistance, and even writing documentation. By handling these tasks, developers can focus more on creative and high-level problem-solving aspects of software development, thus increasing overall productivity.
Enhanced Code Quality: These tools can analyze code in real-time, identify potential bugs, suggest improvements, and ensure adherence to best practices and coding standards. By providing intelligent suggestions and catching errors early in the development process, AI co-pilots contribute to higher-quality code and reduce the likelihood of bugs reaching production.
Knowledge Augmentation and Learning: AI co-pilots often incorporate machine learning models trained on vast amounts of code repositories and developer interactions. This allows them to learn from patterns in code and developer behavior, continuously improving their suggestions and assistance over time. Developers can leverage these tools to learn new programming techniques, explore unfamiliar libraries and stay updated with the latest practices in software development.

Overall, AI co-pilots act as intelligent assistants that streamline development workflows, improve code quality, and empower developers to be more efficient and effective in their work.

AI co-pilots in smart manufacturing are starting to deliver business value

In comparison to what is already everyday business in software development, we can see that AI co-pilots are appearing more frequently in smart manufacturing systems, e.g. as new functionality in modern manufacturing execution systems (MES) or operations-focused low-code frameworks. Although the functionality is not yet on the level of their peers in software development, one can see what is waiting around the next corner. AI co-pilots in smart manufacturing are focusing on the support of the human worker by providing benefits in the following dimensions:

Significantly Improved User Experience: AI co-pilots are enhancing the user experience by offering intuitive interfaces and natural language processing capabilities. Users can interact with MES systems using voice or text, asking for assistance with specific tasks like querying production status, generating ad-hoc performance analysis or generating customized reports. This results in reduced training times and makes it easier for non-technical users to interact with complex systems.
Enhanced Decision-Making Insights: AI co-pilots can process large volumes of manufacturing data (e.g., order status, machine performance, inventory levels) and offer superior capabilities for their insightful visualization including deep drill downs. By that, decision making is backed by faster and more precise analytics than ever before, resulting in more effective root cause identification and instant deviation management.
Overall Improved Operational Performance: By providing growing capabilities to process vast amounts of data for generating insightful data as well as the opportunity to make faster and better decisions, roles on the shop floor from foremen to head of production are enabled to make more efficient use of resources, increase quality of goods produced and to ensure a higher on time delivery to the customer. This guarantees not only overall improved operational performance but also higher margins and financial outcome.

Although this sounds already quite promising, we need to know we are still right at the beginning of AI in smart manufacturing. So, another benefit of adopting AI co-pilots is to get to know this technology early and have time for preparing and mastering it.

Prerequisites to harvest business benefits of AI co-pilots in your company

Facing reality means that most manufacturing companies have not even touched AI in operations at all. Not to mention that still a lot of companies are missing the most basic requirements to step into this game. Therefore, let’s have a look at the major prerequisites to take up this technology and start harvesting its business benefits:

Generating trustworthy data: In data analytics everyone knows about the “shit in, shit out” principle which means that without trustworthy data, basically no meaningful insights can be generated. And this is exactly what applies in AI. Prerequisite no. 1 is to generate as much and as precise data as possible into the backend, including booked times and materials, actual order status, deviations, error codes, downtime reasons etc. This will be the solid ground to build any AI application upon.
Choosing the right partner: Setting up a solid data backbone to realize operations live monitoring, manufacturing analytics and reporting as well as advanced analytics with AI specified insights is no piece of cake. Selecting a partner who is not only able to play buzzword bingo but also has substantial knowledge in delivering hard and reliable functionality is key.
Getting people and organization online: Although ChatGPT is around in the media, interacting in meaningful ways with AI features is still a closed book for many people, especially for the older generation. On the personal level, the ability to formulate meaningful requests and to engineering a targeted prompt become key qualifications for decision makers in operations. Training for people in charge will be mandatory to get used to this new interaction paradigm and to leverage its full potential. On the organizational level, hierarchical decision making needs to be adapted to this completely new level of speed, as it is no longer the insights that are lacking. In this new world, it will be about the willingness to take responsibility and the empowerment to make decisions fast.

AI’s way into smart manufacturing

Having a look at the above-mentioned facts, AI is sneaking into smart manufacturing by supporting the human worker with cognitive augmentation. In detail, this is firstly about processing lager amounts of information in a shorter period. Secondly, it is about enabling better decision making based upon more precise and specific insights. This will result in increased productivity, higher operational performance and - on a company level - an optimized financial outcome.

To get ready for this future, it is highly recommended to do some homework regarding the available data, choose a mate for this tech change carefully and to get people including the entire organization prepared for this transformation to come. As Darwin said, it is not about the survival of the strongest, it is about the survival of the fittest, the one who is the most capable of adapting to changes in its biotope.