Reactive Image Processing: How FDTech Is Bringing Autonomous Driving AI to Industrial Quality Control

Our Goal at FDTech is to further develop AI methods for situation interpretation and fault detection – leveraging our expertise in sensor data analysis and applying it to the challenges of industrial image processing.

The Challenge
Industrial image processing systems capture and analyze visual information to make automated decisions. They play a vital role in quality control across industries such as automotive, semiconductor, food, and packaging — monitoring production processes and ensuring products meet the highest quality standards.

Current image-based quality control systems maintain stable conditions by constructing dedicated inspection setups around the object being tested. While this approach ensures consistent and predictable capture conditions - critical for process stability - it lacks flexibility, as the system is specifically designed for known components and fixed poses.

The core problem is that disturbances such as stray light, varying surface properties, and occlusions caused by loose assemblies can still occur at any time, causing the intended quality control process to fail.

Our Approach
We are researching and developing a method for reactive image processing and its prototypical implementation. The method is designed to analyze disturbances and respond to them autonomously — much like the way a person intuitively examines an object: turning and rotating it in their hand to get a better view from different angles.

Our partners bring complementary expertise to make this vision a reality:

• Kraham Maschinenbau GmbH contributes its experience as a special machine manufacturer, driving efficiency in production quality control.
• Autodeltass develops modular and adaptive control solutions, enabling the flexible automation required for reactive inspection systems.
• Fraunhofer IWU provides deep research expertise in production technology and system integration.
• FDTech advances AI methods for situation interpretation and fault detection, building on our core competencies in automated and autonomous driving.

To validate our research approach, a universal inspection cell with a collaborative robot will be developed and built.

Why This Matters
Robust and flexible quality control is a cornerstone of modern industrial production. By enabling inspection systems to react dynamically to real-world disturbances, we significantly improve the reliability and adaptability of automated quality assurance — contributing to more efficient and resilient production processes across industries.

This project is funded by the Sächsische Aufbaubank (SAB) and runs from February 2026 to January 2028. We are proud to shape this project together with our partners and look forward to contributing to a smarter and more robust future of industrial automation.