On most factory floors, final assembly remains a largely manual process. Whether it’s fitting components on a car chassis or completing bolting tasks, these jobs often require skilled human operators to perform physically demanding, repetitive actions over long shifts. This type of work, while essential, comes with risks: high levels of fatigue, ergonomic strain, and long-term health issues like musculoskeletal disorders.
Automation has long promised relief, but in reality, many existing robotic systems aren’t built to adapt to the complexity and variability of human-centric tasks. They need strict programming, operate in fixed routines, and aren’t always safe or efficient in shared environments. What’s been missing is a system that can genuinely support the worker, not replace them.
That’s the idea behind HEROX.
Developed as part of the ARISE project, HEROX is a collaborative robotics solution designed for real-world assembly environments, where flexibility, safety, and adaptability matter as much as efficiency. It’s a system shaped not only by advanced technology, but by the needs of the people working on the line.
Building Robots that Adapt to People — Not the Other Way Around
At its core, HEROX is built around a Mobile Manipulator (MoMa) that can work safely alongside human operators. The robot assists with heavy or repetitive tasks, such as bolting operations, while the human retains control and decision-making authority. This helps reduce physical strain, improve safety, and open the door for a wider range of workers to engage with robotic systems — including those with limited strength or mobility.
To make that collaboration smooth, the system integrates a few key features:
- 360° Vision and Real-Time Awareness: The robot continuously monitors its environment using a full-surround perception system, allowing it to detect obstacles, track objects, and adapt to dynamic workspaces.
- Voice-Controlled Interaction: Operators can use simple, natural language to guide the robot. There’s no need for custom programming — just clear communication. This is powered by Large Language Models (LLMs), but kept practical in its application.
- XR-Based Visual Guidance: Workers receive real-time visual instructions through XR interfaces, helping them follow procedures more easily and avoid errors — without interrupting their workflow.
- Reactive Control Algorithms: MoMa isn’t locked into a fixed routine. It adjusts its path and behavior in response to human movement and environmental changes, ensuring safety and fluid cooperation.
Designed with Workers in Mind
From the start, HEROX was designed with input from both a technology provider (Neurofai) and an industrial end-user (SAT) to make sure the solution is grounded in real needs. The aim isn’t to showcase futuristic capabilities, it’s to make collaborative robotics useful, reliable, and easy to adopt.
We know the value of human experience in manufacturing, and HEROX is about enabling that experience to thrive. By reducing the physical demands of the job, we can reduce injury risk, shorten onboarding time, and make assembly tasks accessible to a more diverse workforce, including women and individuals with disabilities who may have been excluded by the physical nature of these roles in the past.
HEROX is also committed to open innovation. As part of the ARISE initiative, we’re contributing reusable, open-source modules to help others in the manufacturing community adopt similar approaches. This includes:
- A ROS2-based module for 360° perception and scene understanding
- An LLM-powered voice interface for intuitive task guidance
- A FIWARE-integrated condition monitoring tool for real-time safety alerts and operational insights
- A reactive control layer that can be adapted to different robotic platforms and workspaces
By making these tools modular and open, we hope to lower the barrier for other companies — large or small — to bring safe and effective HRI into their own environments.
The HEROX system in currently being validated in real assembly line conditions with the following targets:
- A 30% improvement in task efficiency
- A 50% reduction in error rates
- And a 20% drop in implementation costs for end users, thanks to modularity and ease of integration
HEROX isn’t about replacing humans with machines. It’s about giving them the tools to do their jobs better, safer, and with more long-term sustainability.
More info here!