Step into the future of robotics

Smarter, Faster, More Adaptive!

TORNADO is revolutionizing robotics with next-gen AI-driven adaptability, real-world impact, and EU-funded innovation. Our autonomous mobile robots (AMRs) navigate complex environments, interact naturally with humans, and manipulate small, soft, and deformable objects like never before.

About

Innovating the Future with AI and Robotics

Quick Highlights

TORNADO in numbers

Next-Gen Robotics

AI-powered adaptability

Real-World Impact

Advanced perception and interaction

About

The vision, the mission, the revolution

Advancing cognitive robotics by integrating:

Cloud-based AI Foundation Models (FMs)

Robots can access, fine-tune, and update pre-trained deep learning models in real time.

Human-Robot Interaction (HRI)

Multi-modal communication through gesture recognition, natural language processing (NLP), and voice commands.

Learning-from-Demonstration (LfD) & Explainable AI (XAI)

Robots acquire new manipulation skills via AR-based demonstrations and self-explanatory AI models.

0

partners from 8 countries, including leading universities, research centers, and industry experts

0

months of cutting-edge research & development (Jan 2025 – June 2028)

0

€ total funding for next-gen AI-driven robotics

0

real-world use cases

Tornado Goals

Tech & Innovation

Develop an Adaptive Cloud Robotics Platform

Enable autonomous mobile robots (AMRs) to navigate dynamic, human-populated environments with real-time AI-driven decision-making.

Enhance Dexterous Manipulation of Small, Soft, and Deformable Objects (SSDs

Implement advanced perception, force control, and machine learning to allow robots to manipulate non-rigid materials with human-like precision.

Leverage Foundation Models for Zero-Shot Generalization

Utilize cloud-hosted deep learning models for on-the-fly adaptation, reducing training time and improving operational efficiency.

Improve Human-Robot Interaction (HRI)

Integrate gesture recognition, natural language processing (NLP), and explainable AI (XAI) to enhance seamless, intuitive collaboration between humans and robots.

Advance Learning-from-Demonstration (LfD) Techniques

Develop AR-assisted skill acquisition methods, enabling robots to learn tasks in real time by observing human demonstrations.

Optimize AI Model Selection & Adaptation

mplement Test-Time Adaptation (TTA) and Out-of-Distribution Detection (OOD) to dynamically select the best AI models for each scenario.

Tech & Innovation

The Technology Behind TORNADO

Cloud Robotics & AI

Robots connected to a pool of pretrained foundation models, adapting to new tasks in real time.


Human-Robot Interaction

Gesture and voice-based communication for seamless integration in workplaces.


Autonomous Skill Learning

Robots learn on-the-fly through Learning-from-Demonstration (LfD) and augmented reality interfaces.

Dynamic Adaptation

Ability to manipulate SSDs of various shapes and materials, adjusting to real-world uncertainty.

Join the Innovation. Explore partnerships:

Follow the TORNADO project on Zenodo to stay updated with our latest publications and research outputs

Zenodo is an open-access repository developed by CERN that allows researchers to share publications, data, and other research outputs with the global scientific community.