Fraunhofer Develops AI-powered Cobot Inspection Workstations for Production

The latest AI-assisted cobot system from the Fraunhofer Institute for Mechatronics Design (IEM) can help reduce product testing and inspection errors by assisting workers in the inspection process.   

 

Manual inspection can be tedious, time-consuming, and error-prone. Image used courtesy of Fraunhofer IEM

 

Manual Visual Inspection

Manual quality testing of components or products in industry can be tedious for workers. Employees have to continuously strain to find problems, which can be exhausting. As a result of the tedium and strain, human errors can occur in the manual part inspection process. Overlooked quality issues can lead to problems during manufacturing or when the product is used, causing headaches for manufacturers and end users. Manual inspection methods alone can slow down production and affect the overall quality of goods.

Fraunhofer IEM has announced the development of a new solution to help mitigate the common problems associated with manual visual inspection. They have collaborated with Diebold Nixdorf, a banking ATM manufacturer, and Verlinked, a software specialist, to create a new solution that combines collaborative robots, AI-based image analysis, and an IoT platform. The new technology hopes to improve inspection by reducing errors and freeing human labor from tedium.

 

Fraunhofer's latest cobot inspection system identifies part problems, allowing workers to focus on fixing them instead of hunting for quality issues. Image courtesy of Fraunhofer IEM

 

A Cobot Workstation 

Cobots are an integral part of Fraunhofer's new system, toting sensors and AI technology to help human workers in inspection. These combined technologies give users a competitive advantage, especially with small batch size inspections, where hard automation can be expensive and difficult to adjust to manufacturing processes with too much flux. Cobots, AI, and human labor create a powerful combination. 

The system, called the it's OWL project CogeP (Cobot-supported test stations for intelligent technical systems), has already begun testing. In the test scenario, employees work alongside a cobot to test and inspect the quality of ATM control panels. To check control panels, a robot arm with sensors moves a camera around the components, workpieces, or products from different angles. AI analyzes the images to assess quality. If defects like loose screws or unstable connections are found, the system alerts workers to fix them. The tested system can be altered with minimal effort to be used in other inspection and testing processes. 

This cobot-enabled inspection is designed to help in quick and precise error correction. It also seeks to help boost productivity and reduce human error because workers are engaged in fixing the problems instead of tediously searching for issues with a workpiece.

The cobot system is designed to help small and medium-sized businesses with production. Image used courtesy of Fraunhofer IEM

 

AI-powered Robotic Inspection

A key feature of Fraunhofer’s inspection solution is the combination of a cobot with an IoT platform for data management. The IoT platform acts as a central point for real-time data, assigning test tasks, saving results, and gathering data from multiple robots and orders. This streamlines testing processes and allows easy adaptation to new product versions without complex and expensive programming costs. With cobot-supported test stations feeding data to the central IoT platform, end users can adjust testing routines for new work without costly physical infrastructure changes.

Fraunhofer has developed its technology to be used for more than just testing products/QC, but also for putting products together, picking orders, and helping out around industrial settings in general. Fraunhofer has integrated its experience in industrial processes and machine learning for analysis to create a versatile new robotic system. They hope to help small and medium-sized businesses add cobot workstations to their factories with greater ease and reduced initial investment expense, helping to save costs even with small batches or unique product lines.