99 susceptibility to errors involved in manually transferring data between tools and databases. The objective is to build a platform that connects multiple applications and sources of information, presents data transparently, analyzes it in real time and facilitates the decision-making process. It could also be used as a basis for implementing further use cases, streamlining communication along the entire supply chain and obtaining forecasts and crafting strategies with greater precision. Efficiency gains in the development phase Brose is already using an AI-based approach today that supports and accelerates the development process for product design engineers: every technical drawing must be checked for a wide range of criteria before it can be approved for follow-up processes. This was formerly an entirely manual and extremely time-consuming process, which is why Brose adopted popular methods from AI research, adapted them and developed its own AI-based solution to increase the efficiency of the technical drawing process. The files are checked using an automatic image and text recognition method and the results are summarized in a system-generated report. This makes it possible to automatically detect and track drawing errors during the final release process, leading to a standardized approach to handling errors that reduces the manual time and effort involved and results in an annual cost savings in the five-digit range. Intelligent production technologies ensure quality levels Artificial intelligence is also a promising technology for production. Brose optimizes its production lines on the basis of intelligent monitoring, data analysis and inspection systems to significantly increase the efficiency of its quality assurance systems. SMT lines in electronics manufacturing are a perfect example of this. SMT stands for “Surface Mounted Technology”. It describes an assembly method during which miniaturized electronic components are placed on the surface of a PCB. Brose inspects around 300,000 components per hour with the help of an automated optical inspection system in an electronics production line. Previously, image processing was implemented without artificial intelligence and only achieved accuracy levels between 85 and 95 percent. Ambiguous results had to be manually reinspected and classified into true and pseudo defects. Reasons for pseudo defects can include issues related to component tolerances, solder joint quality, reflections or extraneous light influences, for instance. Thanks to the use of machine learning, the AI models developed in-house are able to detect these pseudo defects. These models are clearly successful: initial output was increased to up to 99 percent and manual inspection effort was decreased by 80 percent. Another area of application for artificial intelligence is the product itself. AI is suitable for integration in existing Brose components and systems and for the development of innovative approaches for entirely new mobility experiences. E-Mail info@brose.com Brose www.brose.com Image: © Brose An AI-based AOI process on SMT lines in electronics manufacturing significantly increased initial output to up to 99 percent. AI models developed in-house automatically detect pseudo defects, which means they no longer need to be manually reclassified. Image: © Brose Image: © Brose Contact Martin Tupy Head of Applied Artificial Intelligence Brose Group
RkJQdWJsaXNoZXIy MjUzMzQ=