Fraunhofer NeurOSmart Platform Enables Safe and Energy Efficient Human Robot Collaboration in Industrial Production

© Fraunhofer ISIT
The movable MEMS mirrors project the LIDAR laser across the entire work area. The reflection of the laser pulses creates a high-resolution 3D image of the work area shared by humans and robots.

(IN BRIEF) Fraunhofer has developed the NeurOSmart technology platform to enable safe and intelligent collaboration between humans and robots in industrial environments. The system combines LIDAR-based sensor technology, AI-supported data processing and neuromorphic chips that mimic brain-like decision-making to provide fast and energy-efficient robot control. A high-resolution 3D sensor system monitors shared workspaces while embedded AI processes data directly at the sensor level, reducing latency and power consumption. Neuromorphic accelerator chips allow real-time decision-making within milliseconds, enabling robots to safely slow down or stop when people approach. Developed by several Fraunhofer institutes, the modular platform can be adapted to various industrial applications and may also be used in areas such as drones and agricultural sensing systems where energy efficiency is critical.

(PRESS RELEASE) MUNICH, 3-Mar-2026 — /EuropaWire/ — Fraunhofer has developed the NeurOSmart technology platform to enable intelligent and safe collaboration between humans and robots in industrial production. By combining advanced sensor technology, artificial intelligence and neuromorphic chips inspired by the human brain, the system allows robots to operate alongside people with greater awareness, speed and energy efficiency.

Modern manufacturing increasingly depends on robotics that can do more than repeat simple tasks. For robots to assist human workers effectively, they must be capable of executing complex movements, responding quickly to instructions and maintaining safe distances at all times. The NeurOSmart flagship project addresses these requirements by integrating sensing, data processing and robot control into a unified platform designed specifically for collaborative production environments.

© Fraunhofer IWU
With intelligent sensor technology and AI-supported data processing, NeurOSmart technology enables close collaboration between humans and robots.

The project is coordinated by the Fraunhofer Institute for Silicon Technology ISIT and involves several Fraunhofer institutes, including the Fraunhofer Institute for Photonic Microsystems IPMS, the Fraunhofer Institute for Microelectronic Circuits and Systems IMS, the Fraunhofer Institute for Machine Tools and Forming Technology IWU and the Fraunhofer Institute for Intelligent Analysis and Information Systems IAIS. The partners have jointly developed and integrated the individual technologies into a complete system that allows humans and machines to work together safely and efficiently.

At the core of the platform is a sensor system that continuously monitors the shared workspace of humans and robots from above. The system uses LIDAR technology, which emits short near-infrared laser pulses and calculates distances based on reflected signals to create a detailed three-dimensional representation of the environment. Movable MEMS mirrors distribute the laser beam across the entire work area, producing a high-resolution 3D image of the shared workspace.

The MEMS mirrors have been optimized for both performance and energy efficiency. They are manufactured using piezoelectric aluminum scandium nitride with a thickness of only one micrometer, allowing precise and rapid movement while minimizing power consumption. This sensor technology provides continuous monitoring of the robot cell and forms the basis for safe human-robot interaction.

A defining feature of NeurOSmart is that signal processing takes place directly inside the sensor system. The large volume of data generated by the wide-field sensor is first preprocessed locally. Artificial intelligence algorithms developed by Fraunhofer IMS analyze the incoming signals and identify areas of particular relevance. This allows the sensor to focus on important regions for further analysis, reducing both data transmission requirements and energy consumption.

The final evaluation of sensor data and robot control decisions also occur directly within the system through neuromorphic computing technology. Researchers at Fraunhofer IPMS have developed a specialized accelerator chip composed of many interconnected computing units arranged in a matrix across a wafer. Each unit operates independently, functioning similarly to neurons in the human brain and enabling distributed decision-making. This architecture allows complex calculations to be performed quickly while consuming far less energy than conventional computing approaches.

Fraunhofer IAIS has created the AI models used for interpreting sensor data and controlling robot behavior. The integrated system can respond within milliseconds, enabling heavy industrial robots to slow down or stop automatically when a person enters the safety zone. Virtual simulations of complete robot cells have been used to train the system, including hazardous scenarios that cannot be safely reproduced in real environments.

The neuromorphic approach offers a significant advance in hardware architectures for demanding AI applications by combining high processing speed with low energy consumption. The NeurOSmart platform is designed to make industrial production more flexible and efficient while improving workplace safety and usability.

The NeurOSmart technologies are structured as a standardized platform that can be adapted for different industrial applications. Manufacturers can develop customized solutions based on the modular components to suit their specific production environments. Beyond factory automation, the energy-efficient sensing and processing technologies could also be applied in mobile systems such as drones or agricultural sensor platforms, where long battery life is essential.

Media Contact:

Claudia Buschmann
Head of PR and Communications
Fraunhofer Institute for Silicon Technology ISIT
Fraunhoferstraße 1
25524 Itzehoe
Phone +49 4821 17-3627
claudia.buschmann@isit.fraunhofer.de

SOURCE: Fraunhofer-Gesellschaft