AI-Driven Inspection: Future of Quality Control

Abstract digital landscape of glowing blue data points forming wave-like structures on a dark background.

As industries and production processes continue to evolve, the role of artificial intelligence in inspection processes is becoming increasingly important. But do you know what possibilities are already available today and how AI can support you in your inspection process? We asked Christian Wojek, Head of AI at ZEISS Industrial Quality Solutions, to explore how AI is shaping industrial metrology and the future of quality control.

Christian, what are the key drivers for using AI in inspection processes?

Well, as industries transform, we’re really seeing a shift in quality control from dimensional metrology applications in the measurement room centered around a metrology expert to automated inspection and metrology that are integrated into production processes. This change is all about increasing efficiency and embracing automation. The key drivers for adopting AI include the need to boost productivity, reduce errors, and tackle challenges like price pressure and the shortage of skilled workers. Customers are looking for high quality, fully automated, highly accurate, and flexible solutions at the same time. With the extensive industry experience of ZEISS we’re in a great position to be a trusted partner, delivering comprehensive, user-friendly AI solutions tailored to meet the diverse needs of various industries.

What AI technologies are currently being utilized in inspection processes?

At ZEISS, we’ve integrated AI into various applications for years. We began with inline CT in casting and microscopy, focusing on layer thickness measurement, particle detection for cleanliness, and grain analysis. More recently, we’ve expanded our offerings to include AI-based inspection options for electron microscopy applications such as metallography. For CT and X-ray applications, we enable AI-driven inspections, such as noise reduction in X-ray imaging and improved capabilities for weld inspection and defect detection in battery cells. AI also plays a crucial role in identifying issues like porosity (in castings, for example) and supports non-destructive testing (NDT).

How does AI enhance inspections compared to traditional methods, and where is it applied?

AI significantly improves inspections by managing challenging imaging conditions, such as the reduction of noise and artifacts in CT images, which leads to fewer false positives and more reliable results. It can detect defects that traditional software might miss, performing at a human recognition level. Implementation time has drastically decreased, enabling quick AI model creation after scanning samples. AI is used in both measuring rooms, where ease of use is key, and on shop floors, where speed and performance are prioritized, whereby it’s always ensured that tools remain user-friendly for quality control personnel.

What options are available from ZEISS for customers interested in AI solutions?

First of all: our AI solutions are easy to use for everyone. You don’t need to be an AI expert to use it. To develop customized models, customers can collaborate with our ZEISS AI application experts. We’ve already created over 100 AI models for microscopy and CT applications and offer pre-trained models for common use cases. Customers can also train their own deep-learning models with microscopy data, and this capability will extend to other technologies like CT data in the future. Of course, the execution of AI models seamlessly integrates into ZEISS software, enhancing image analysis efficiency.

Various ZEISS microscopy and imaging systems are displayed on separate white platforms, connected digitally.

AI-driven analysis for microscopy

With our microscopy software ZEISS ZEN core, you can acquire images, analyze samples, and integrate all your data into a unified ecosystem for connected microscopy. AI-based algorithms can be applied effortlessly, with no programming required, throughout the entire workflow from imaging to analysis. With the latest release of ZEISS ZEN core, it is now available for electron microscopy, giving you access to the entire microscopy portfolio for a connected research environment.

Learn more

Close-up image of a star-shaped metallic component with a central black circular opening.

Can you provide examples where AI has significantly improved inspection processes in different applications?

Sure. For instance, Smith & Nephew, a medical tech company, uses our AI-driven software to automate the microscopic inspection of implant coatings, reducing the measurement time from 45–60 minutes to just 5–7 minutes – a tenfold increase in efficiency. And Festo, a leader in industrial automation, employs our AI-supported automated defect recognition for porosity analysis with CT data. How? At ZEISS, we trained AI models to identify patterns in casting parts, allowing for automated segmentation and classification of defects, which helps Festo to quantify and judge subtle irregularities more effectively.

What exciting trends in AI do you see today for inspection? How will future advancements impact inspection processes?

AI is no longer a concept of the future; it’s actively being utilized across various industries today. We see a shift towards end-to-end solutions that enable holistic process optimization. This involves connecting process data with quality data to enable closed-loop control. AI can play a crucial role here. And I think in the long term we’re approaching an era of autonomous inspection systems in “dark factories,” where minimal human intervention is needed. In these environments, machines, robots, and AI systems autonomously handle tasks like production, assembly, and quality inspection, streamlining operations and enhancing efficiency. There are exciting times ahead.

About Dr. Christian Wojek
Head of Artificial Intelligence, ZEISS Industrial Quality Solutions

Christian Wojek drives the adoption of AI at ZEISS Industrial Quality Solutions, enhancing product development and internal processes. He actively supports multiple teams in leveraging AI to drive product innovation and improve workflows. Previously, he led efforts in developing cutting-edge computer vision solutions for X-ray and CT imaging. Passionate about translating the latest advancements in machine learning into practical applications, Christian is dedicated to delivering impactful solutions that empower customers in their daily operations.

Ensuring quality in AI data centers

The race to integrate AI and machine learning into data centers is moving at lightning speed. Ensuring high-speed data communication by using high-performance hardware components is key. Discover how Quanta Computer achieves top quality.

Read now