"Identifying components, comparing work steps and correcting errors - that is what Smart Klaus is already able to do today. Updating, expanding and optimizing his own database independently - that is what Smart Klaus will learn in 2020" says Wolfgang Mahanty, Managing Director of Optimum datamangement solution GmbH. The Karlsruhe-based company specializes in intelligent image processing that combines database management and camera technology. Optimum has been able to gather more than 25 years of experience, which today is completely used for the further development of Smart Klaus. The system controls and supports complex manual manufacturing processes. It guides employees, accompanies them through their tasks and ensures quality at every step of the process.
When developers speak of artificial intelligence, they mean the ability of programs, machines and computers to learn new things on their own. In addition, these systems should be able to actively convert the acquired knowledge into actions. The core of the technology is a neural network that supplies a range of data. These are ordered, evaluated and abstracted by the program. The organized information serves as a basis for further decisions.
An artificial intelligence evaluates these data, compares them with each other and recognizes matches. Using previously learned systematics, it is then able to correct errors in its own database. What was previously primarily of interest for research is now also finding its way into industry thanks to faster computers and better networking.
To perform his tasks, Smart Klaus already has a number of sensors that determine parameters such as contour, dimensions and weight. There are also three neural networks that are used to recognize texts, colors and textures. However, their ability to independently generate further information and add it to their own database can still be expanded.
Optimum's plan is to make the camera-based assistance system not only smarter, but intelligent. The magic word is Deep Learning. In order to ensure greater accuracy with articles such as cables, fabrics and covers, which are a challenge due to their variable properties, the clever Klaus is to teach himself how the articles look in their various variations.
"These skills are particularly interesting for camera-supported product recognition. With regular use, an artificial intelligence increases the accuracy and also the speed of the system considerably." Wolfgang Mahanty is convinced. "Artificial intelligence is getting smarter and more efficient, making it the perfect support for tomorrow's workplace."