Share

Smart interactive robots serving humans

CEA-List’s smart robotics research program aims to design and develop a new generation of robots that offer unparalleled performance, are more interactive, and are better at learning new tasks. To tackle this issue, we are working to bring the necessary technologies to maturity, of course. But we are also looking at smart interactive robots as systems. Specifically, we are outlining the contours of an architecture that will enable the integration of heterogeneous technologies into increasingly sophisticated robots. Our companion robot demonstrator is the first incarnation of this approach.

At CEA-List, our smart robotics research program is driving the emergence of a new generation of robots. We are developing innovative mechatronic solutions and integrating the latest advances in embedded systems and software to make tomorrow’s robots—whether they are fixed or mobile—more powerful than ever. These robots will possess intrinsic learning capabilities that will make them smarter. They will be capable of quickly mastering complex new tasks and performing them autonomously.

New interfaces will facilitate efficient and natural interaction and collaboration with humans, machines, and other robots. Connectivity and communications capabilities will enable seamless integration of robots into their digital environments. As a result, these robots will be more versatile, easier to use, and more adaptable to new use cases.

These robots will be companion workers for humans, helping respond to the grand challenges of flexibility and agility of the Factory of the Future. They will contribute significantly to the performance of companies of all sizes. And production is just the beginning. This new generation of robots will be just as at home in logistics, health, and the circular economy as they are on a factory floor.

Two research areas

CEA-List’s smart robotics research program focuses on two research areas in pursuit of the following objectives:

  • Develop digital twins of robotic systems. These advanced real-time simulation tools will simplify the design process and optimize the integration and operation of robots in their operating environments.
  • Give robots enhanced perception capabilities, especially regarding vision and force; improve the performance of key robotic functions such as moving around, grasping, two-handed manipulation, and collaborating with human operators.
  • Make robots context-aware and able to make decisions in full autonomy in order to adapt optimally to their operating environments.
  • Enable more efficient and natural collaboration with human operators and with other robots through intuitive multi-modal human-machine interfaces (HMIs).
  • Give robots the ability to learn, either by observing human operators complete tasks or with AI learning techniques, so that they can rapidly adapt to new tasks and situations.
  • Enable robots to share the knowledge and skills acquired via a unified knowledge representation.
Research area 1

Towards an open and modular robot architecture

To simplify and accelerate the adoption of intelligent robotics in industrial and other settings, CEA-List is developing a system architecture that will allow optimal interaction between all of the technologies used in smart robots and make it easier to adapt robotic systems to different use cases. Another objective is to make smart robots safer, more autonomous, and more versatile.

Read Architectures and software for greater robustness, adaptability, and autonomy

Research area 2

Better components for better robotic systems

Giving robots advanced perception and action capabilities is one of the challenges inherent to smart robotics. These capabilities are necessary in order for robots to be able to efficiently carry out a wide variety of complex tasks in changing environments and for them to be able to share space and work with human operators.

At CEA-List, we feel that the best way to achieve this is through a system-level approach to developing a high-performance, scalable, and safe robotics platform. To this end, we are conducting unified and coordinated research and development in mechatronics, computer vision, artificial intelligence, human-machine interfaces, and communications networks addressing both the robots themselves and perirobotic functions. Digital twin technology is one of the key tools supporting our work.

Read Robotic systems: Integrating and linking technologies for smart robots

The CEA-List companion robot demonstrator

CEA-List’s companion robot demonstrator brings the institute’s early smart robotics research to life. The system can assemble, either independently or with the help of an operator, a set of around ten mechanical parts placed randomly in a bin. To complete the task, the robot must be able to coordinate several complex functions:

  • Planning, which entails recognizing and localizing the parts in the bin, determining access paths, picking up and assembling the parts, and calling the human operator if necessary.
  • Vision, in a context where the parts are small, similar to each other, and shiny.
  • Grasping and assembly of different-sized parts with smooth surfaces that fit very tightly with each other and that need to be screwed together.
  • Human-machine interface to monitor the procedure, call the operator, analyze the work being completed, and manage rework.

 

In the future, the companion robot’s flexible architecture will allow it to learn and carry out new tasks, including multiple assemblies with more complex and varied parts (assemblies combining rigid and flexible parts, for example), and assembly scenarios involving subassemblies.

In the longer term, our goal will be to adapt the companion robot to use cases provided by CEA-List R&D partners like Areva, Stellantis, and Technip, for example.

Technology providers like Haption, Isybot, and Staübli, also CEA-List partners, could also integrate these technologies into their solutions.

Strong expertise in the many fields that make up intelligent interactive robotics

CEA-List conducts its own research in many of the advanced technologies used in robotics. These include interactive robotics, control engineering, computer vision, deep learning, natural language processing, cybersecurity, software engineering and integration, and software and functional certification. Having such a broad range of expertise available right nearby helps keep the robotics program’s projects moving forward at a rapid pace.

European programs

CEA-List has been a major contributor to smart robotics research and development for more than a decade.

The institute coordinates the following European projects:

  • MERGING (Manipulation Enhancement through Robotic Guidance and Intelligent Novel Grippers)
  • RobMoSys (Composable Models and Software for Robotics Systems)

CEA-List is currently or has been a partner in these ongoing and completed European projects:

  • AiBle
  • COVR (Being safe around collaborative and versatile robots in shared spaces).
  • ROMANS (Robotic Manipulation for Nuclear Sort and Segregation
  • TRACEBOT (Traceable Robotic Handling of Sterile Medical Products)
  • ESBF 2 (Autonomous navigation of a RATP bus)
  • ESPRIT Stabilizing control of a convoy of cars

The institute is also a member of the following robotics networks of excellence:

  • RIMA (Robotics for Inspection and Maintenance)
  • AGROBOFOOD (Towards a European network and effective adoption of robotics technologies)
  • TERRINET (The European Robotics Research Infrastructure Network)
  • HORSE (Smart integrated Robotics system for SMEs controlled by Internet of Things based on dynamic manufacturing processes)
  • ECHORD ++ (European Coordination Hub for Open Robotics Development)
  • HERO (Towards a European network and effective adoption of robotics technologies)
  • EU Robotics
  • ESMERA (European SME Robotics Application

See also

Technology platforms

SMART interactive robotics platform

Improve robots’ capabilities and develop new ways of interaction with humans.
Read more