Person and Object Detection

Modern production systems must reliably detect people and respond appropriately, even when people are obscured, making unusual movements, or when unknown objects are present. Fraunhofer IKS develops modular safety solutions for this purpose. These solutions enable companies to safely and reliably deploy and test AI-enabled perception even in dynamic production environments.

Safety as enabler for industrial AI adoption

Fraunhofer IKS’s approaches to reliable person detection offer:

• Robust detection
• Context-aware interpretation
• Safe real-time responses
• Demonstrable assurance of AI systems

This allows companies to take advantage of the efficiency gains through AI; while ensuring safe, validatable, and trustworthy deployment in industrial production environments.

Safety thus becomes the bridge between innovation and industrial adoption. This bridge is particularly important for Physical AI – AI-enabled machines, robots, and autonomous systems that share workspaces with people. The future of industrial AI depends not only on what these systems can achieve, but also on how reliably and transparently they can be tested, validated, and trusted in real-world operation.

Fraunhofer IKS is conducting research on various approaches and methods that complement one another but can also be used independently of one another:

Safety Net

ODD Assist Agent

PROWL: Zero-Shot Detection

Dynamic Safety Logic

Contact us now

The Fraunhofer IKS Safety Net describes our modular approach to AI-based perception systems. According to the demands of the actual scenario, it flexibly combines different detection methods, including object and body-part detection, depth and motion analysis, and out-of-distribution detection, together with a precisely calibrated multi-camera system designed for optimal coverage of the production environment. This redundant architecture makes it possible to reliably identify edge cases in which conventional person detection models often fail, such as unusual poses, partial occlusions, unknown objects, or complex interaction scenarios.

The Fraunhofer IKS Safety Net therefore provides the foundation for demonstrably validating and verifying safe, flexible AI systems in dynamic production environments. It integrates perception, contextual assessment, uncertainty analysis, and safety logic throughout the entire lifecycle—from development and integration to operation.

To assess safe operating conditions, we use our ODD Assist Agent, which enables LLM-based analysis of the Operational Design Domain (ODD). This makes safety-relevant contextual information explainable, auditable, and usable in industrial applications.

You can find out more about the ODD Assist agent in the following blog acticle.

How to ensure safe operating conditions around robots?- ODD Agent for Industrial Automation using LLMs