The Siemens factory Electronics Works Amberg of Factory Automation is building different products out of the SIMATIC S7 family. This includes, besides R&D, the assembly, soldering, housing and packaging of the end products ready for shipping to customers.
Objectives: Automated visual inspection solution to reduce manual effort and costImplement a neural network directly at the field level Cost effective solution with low initial investment in AI hardware and AI standard softwareBuild-up of a closed-loop environment for data collection, AI modelling, training, deployment, inference and monitoring based on Siemens products for scaling of AI use cases

Low-cost pin in paste solder joint inspection using computer vision based on a deep learning algorithm:Visual inspection of soldered PCB boards to identify quality relevant solder spots automatically. The goal of the project is to solve the use case and at the same time create the infrastructure for scaling AI use cases in production based on Siemens Industrial Edge.
AI training environment based on Siemens portfolio on cloud and packaging via AI SDK, deployment via AI Model Manager to Industrial Edge devices running by AI Inference Server. The complete infrastructure is based on Siemens hardware and Siemens apps or for cloud environment (AWS) and 3rd party industrial cameras.