Nicolas Mansard, coordinator of the MEMMO project, winner of Stars of Europe
Created in 2013, Stars of Europe rewards coordinators of European collaborative research projects. On December 6, Sylvie Retailleau, Minister of Higher Education and Research presented the trophy to the twelve winners in a ceremony at the Quai Branly museum. Among them, Nicolas Mansard, CNRS researcher in robotics at LAAS-CNRS, holder of the ANITI chair ” Artificial and natural motion”, awarded for the coordination of the MEMMO (memory of motion) project.
Funded by the Horizon 2020 program for four years, MEMMO (Memory of Motion) is a collaborative project initiated in 2018 that brought together a consortium of 10 European partners for a budget of 4 million euros: the LAAS-CNRS ( France) , IDIAP (Switzerland), University of Edinburgh (UK), Max-Planck Institute (Germany), University of Oxford (UK), Trento University (IT ), PAL-Robotics (Spain), Wandercraft (France), Airbus ( France) , Costain (UK) and APAJH (France).
Nicolas Mansard, robotics researcher in the Gepetto team of LAAS-CNRS (Movement of Anthropomorphic Systems), CNRS bronze medalist, project coordinator, announced at the ceremony:
“I want to thank the people who helped me coordinate this project. This is a project set up by a consortium of young researchers. It is a great honor for me to be chosen to coordinate this project”.
Then he added:
“We wanted to prove that it is possible to generate complex movements for arbitrary robots with arms and legs interacting with a dynamic environment in real time”.
Memorizing optimal movements
Calculating robot movements is complicated and more so for robots with arms and legs, which move in an unstructured environment.
For this project, the consortium brought together a team of numerical optimization experts, machine learningrobot control and design.
Nicolas Mansard explains:
“For a walking robot to react to a situation in real time, it must solve a numerical problem with 10,000 variables in a millisecond, which is beyond the scope of what artificial intelligence can do today. We invented Memory of Motion to meet this big challenge”.
The project team generated a huge amount of pre-calculated optimal movements offline and compressed it into a database of possible reactions called Memory of Motion.
Nicolas Mansard commented:
“We use the best motion planners available to minimize exploration time [de la base de données] and improve the quality of the data generated, and we use machine learning to encode it into motion memory, which takes up less storage. Then, we efficiently adapt a candidate’s memory movement to similar situations that have not been explicitly explored. This is called “generalization”.
When it moves, the robot recognizes a new situation thanks to its sensors in real time. It then selects the appropriate reaction from its memory and optimizes it using its predictive capabilities.
Nicolas Mansard added:
“Online, we use this motion memory to guide an ‘optimization solver’ that makes the final decision on how the robot should behave to maintain balance, walk, manipulate tools and other things.”.
Three demonstrators proved the proof of concept
The team developed three demonstrators:
- A humanoid robot to perform handling tasks for aircraft assembly one in a factory of the future with consortium partner Airbus in Toulouse;
- A walking exoskeleton was paired with a paraplegic patient in a rehabilitation center under medical supervision, in collaboration with the APAJH federation, an association for adults and youth with disabilities. A medical center will also experiment with it.
- A quadruped robot capable of walking through a tunnel being excavated or through buildings to be demolished, in collaboration with Costain, the British developer of the Channel Tunnel. It was also tested with an existing commercial industrial inspection robot from the Swiss company ANYbotics.
The same motion generator was used to build three different demonstrators thanks to numerical optimization and machine learning, proof of concept was performed for all three, making it possible to imagine future applications.
Currently the humanoid robot exists only in the laboratory, the quadruped works but is not sold, because for the exoskeleton it is a tool used in rehabilitation centers.
Nicolas Mansard says:
“Now the exoskeleton allows patients to work on their rehabilitation in a hospital center, but in the long term, it can replace the wheelchair.”