Automation

The AI-driven, self-learning digital twin continuously adapts to real system behavior, ensuring an optimal representation of the production process. A comprehensive semantic model serves as the foundation for advanced artificial intelligence (AI) approaches. Insights derived from AI methods are integrated into this model, enhancing the interpretability and explainability of AI systems. Techniques from the field of eXplainable AI (XAI) facilitate the automated description of AI models and their findings, as well as the development of self-explanatory models.

In modern production companies, a heterogeneous IT landscape often complicates day-to-day work. A promising antidote is the use of intelligent agents, which use generative AI for routine tasks and can therefore increase efficiency. Whether these intelligent systems can be successfully integrated into existing networks determines whether the flow of information can be improved and manual effort reduced.

The planning and implementation of automated warehouses is characterized by high investments and risks. The FMEA (Failure Mode and Effects Analysis) currently used to reduce risks requires a great deal of effort to conduct, as it has deficits in terms of design and implementation support. These deficits include a predominant focus on the process view without linking this to the design FMEA for automation objects, an insufficient structure for the use of similar repetitive processes and technologies, a lack of automated, parameterized generation of activities, failures and causes, and a lack of integrated test scenario derivation. These deficits lead to unrecognized failures and increase the effort required to carry out the FMEA and develop test scenarios. In this article, we present a generic FMEA model which, among other things, is able to access extensive practical data in the form of knowledge bases and thus resolve the aforementioned deficits.

The complex dynamics of block warehouses pose major challenges to the manual stocktaking process. Frequent relocation of pallets, crates or pallet cages without fixed storage locations leads to a time-consuming and error-prone inventory process, wherein goods often have to be searched for and damages due to improper storage can occur. The use of (semi-)autonomous drones offers a promising solution to enable automated stocktaking, especially if these are appropriately equipped for optical goods detection.

Motion-Mining® is a technology that uses motion sensors and pattern recognition to enable automated process mapping and analysis of manual work. This article evaluates the advantages and limitations of its use in manufacturing and logistics processes. To this end, Motion-Mining® is compared with traditional lean management tools used to analyze manual activities. Experiences derived from four use cases provide decision support for selecting the appropriate method for a specific use case.

The Enterprise-Control System Integration of the operational applications is described in IEC-62264 and also referred to as the automation pyramid. This integration model is built on the MRP-II model developed in the 1980s. This model was groundbreaking for its time and still forms the basis of operational IT systems today. According to this concept, operational applications are run through hierarchically-sequentially (waterfall principle), which results in disadvantages such as: many interfaces, time delays, data loss, inconsistencies, etc. This sequential model neither meets the current requirements nor the informational and technical possibilities of Industry 4.0. It can be replaced by the concept of the digital control twin, which has corresponding effects on the automation pyramid.

Routing flexibility enables a robust, resilient design of production. However, in highly available, decentralized controlled production systems with cyclic material flow, it can reduce efficiency due to undesired deterministic-chaotic behavior. An automated method for measuring such behavior is presented. It is tested with a double conveyor belt laboratory system. An embedded system simplifies data acquisition. Results indicate that the method is usable for manual and automatic production systems. It has the potential to recognize modeling deficiencies in Industry 4.0 control with IEC 61499. (Only in German)