Safety Net – Dependable Person Detection in Industrial Environments

Are you ready to transform safety in your industrial workspace? Our approach empowers organizations to implement reliable person detection systems, ensuring a secure environment for employees while optimizing productivity. At Fraunhofer IKS, we specialize in integrating Artificial Intelligence (AI) into safety-critical applications, making the collaboration between humans and autonomous machines seamless.

Challenge: Ensuring Safety with Autonomous Machines

As industrial environments adopt more autonomous systems, ensuring the safety of human personnel alongside these machines becomes paramount. Conventional safety measures can hinder operational efficiency, leaving a gap in the ability to maintain safety without compromising productivity. Our framework addresses these challenges by developing a camera-based person detection system, paving the way for safe and efficient operation in shared workspaces.

Solution: Iterative Development and Safety Assurance

Rather than relying on conventional safety measures, our approach focuses on an iterative development process guided by the ML Safety Lifecycle, as defined in the recent standard ISO/PAS 8800 from the automotive domain. We are adopting this approach for industrial applications, such as production facilities and warehouses. By continuously refining safety requirements, collecting relevant data, and evaluating performance, we ensure that our person detection systems operate safely under real-world conditions. This method not only identifies potential hazards but also implements robust mitigation strategies, enhancing overall safety. In a joint project with the Fraunhofer IGCV, we demonstrate in a lean setup how this camera-based approach can be implemented as an assistance system in addition to conventional safety measures like light barriers.

Key Benefits:

• Safe Person Detection: Ensure worker safety with a camera-based approach and reliable detection algorithms that can accurately distinguish between humans and machines.
• Increased Efficiency: Maximize productivity by reducing unnecessary operational restrictions typically imposed by traditional safety measures.
• Adaptive and Flexible: Our systems can be easily tailored to meet the specific safety needs of various industrial environments, allowing for quick adjustments as operational requirements evolve.
• Confidence in Collaboration: Foster trust between human workers and autonomous machines through proven safety mechanisms and transparent processes.

Insight Background

At the forefront of innovation, we aim to create a safer and more efficient future in industrial environments. Our research addresses the critical question: "How can AI-based person detection systems reliably coexist with autonomous machines?" By providing a structured approach to safety assurance, we empower organizations to navigate the complexities of human-robot collaboration effectively.

Our research focuses on trustworthy artificial intelligence, safety assurance, and resilient software systems. Join us in revolutionizing the way AI is integrated into safety-critical environments, making person detection accessible to all industrial sectors.

In-depth information

• Person and Object Detection