12.11.2021

Machine Learning - what you should know about it

Whether in the production of goods, autonomously driving cars or improving our energy supply: the use of artificial intelligence or machine learning (ML) is on the rise, and hardly any branch of industry in Germany will be able to do without it in the future. But what exactly is behind these terms?

Dieser Beitrag ist auch verfügbar auf: Deutsch (German)

What is artificial intelligence?

When machines or systems are enabled to work "intelligently", we speak of artificial intelligence, a branch of computer science. However, it does not specify what is meant by "intelligent" or how this goal is to be implemented or achieved technically.

What you should know about Machine Learning

Machine Learning (ML) refers to a concrete intelligent system, which is also understood in expert circles as a "key technology of artificial intelligence". The use of ML aims to generate knowledge through experience and thus enable the system to improve itself. For this purpose, large amounts of data are collected and evaluated, so that patterns and regularities can be specifically looked for. These recognized rules and a model developed from them can then be applied to as yet unknown data in the next step. Which in turn can help, for example, in improving the predictability of future events.

Asimple example of the use of ML in manufacturing is determining the optimal time to replace a gearbox. If, for example, data is collected on the running noise and vibrations of the gearbox during its service life up to the point of failure, it is possible to predict when failure will occur here for any further gearbox of this type by monitoring the same parameters.

Since ML is also largely based on methods from statistics, the same applies here: The larger the data basis, the more accurate results can be achieved. "Learning machines" do not always have to be of a physical nature, by the way; so-called chatbots, which we frequently encounter while surfing the Internet on various help pages, are also based on this principle, for example.

Technical and social challenge

ML seems to be the solution to many problems - However, despite all the euphoria, it is important to take into account the challenges must be kept in mind.

Thus, for the successful use of the technology first of all, practical application scenarios, so-called "use cases" have to found . The opportunities offered by the technology are often overestimated or underestimated. In the further course, problems often occur with the Data quality. Which data are suitable and how do I evaluate them correctly ? Meticulous preparation is essential if ML is to be successfully integrated into everyday work.

In addition, the "social" challenges to the use of ML should not be underestimated. Many of us are familiar with the horror scenario from science fiction movies: the overpowering machines have seized world domination and are now oppressing humanity. As abstruse as this idea may sound to many of us, numerous people, especially those outside the field, still view "intelligent machines" with skepticism. But in addition to these fantasies, some people also have very concrete fears: The fear of losing their job, of being "replaced" by the use of machines.

This is where education and highlighting the myriad of opportunities that arise from the use of ML, which ultimately can benefit us all, can help. For example, the use of ML can create many new opportunities for jobs.

What's next?

As already shown, AI and ML hold great profit potentials and challenges. Almost 60 percent of German companies already use at least one machine learning application. More than half of the companies rely on the help of external specialists for implementation.

We, pragmatic industries GmbH, in cooperation with pragmatic minds GmbH, support you on your way into this new age. We help our customers with the selection and implementation of AI projects.

Did you like the article? Then share it with others
The OEE - the king indicator among the production ratios
03.11.2021
Noch nicht genug?
Besuchen Sie uns doch hier mal

Browse further

Selection

Selection
  • We, the pi (14)
  • Knowledge (7)
  • Events (5)
  • Open Source (2)
  • Press (2)

Search

Search
Anna - the organization ace
18.11.2021
Predictive maintenance - a core component of Industry 4.0
18.11.2021
Marco - master of data analysis
18.11.2021
Julian - environmental sow?
18.11.2021
1 2 3 6

Load

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram