Innovation and Robotic
Inria at Innorobo 2014
Several teams and Inria reseachers are taking part in Innorobo 2014, the International Summit on Robotics taking place in Lyon on 18-20 March.
Poppy, humanoid robotic platform
© Inria / Photo H. Raguet
Poppy, humanoid robotic platform - © Inria / Photo H. Raguet
The Flowers team investigates the mechanisms that enable robots and humans to learn new motor, interactive and social skills.
It presents the humanoid robot Poppy it has designed leading research on walk mechanisms learning in humans. Its originality is to be both open-source and based on the 3D printing, which allows to quickly explore new forms of its body. Flexible and easy to use, Poppy is an experimental platform dedicated to the world of education and vocational training or association (eg FabLabs ) for learning 3D printing technologies for the design and rapid prototyping.
The team also presents the collaborative European research project for robotics in industry "Semi- Autonomous 3rd Hand"it coordinates. This project aims to achieve critical scientific and technological advances in the field of robotics where robots can work alongside workers to work intuitively and safely on the assembly lines.
The Lagadic team bilocated Rennes and Sophia Antipolis, is specialized in robotic vision: visual control, real-time tracking in image sequences, SLAM, 3D mapping. Its goal is to develop strategies for perception and action from images for applications in the fields of robotics and intelligent vehicles (location, navigation, manipulation, medical robotics), as well as augmented reality. The team partner Équipex Robotex within the theme HRNI (humanoid robotics and natural interactions).
The team will present its work in two areas:
- Automatic Driving: 3D mapping and autonomous navigation in an urban environment.
- Assistance to persons with reduced mobility.
TurtleBot and WifiBot, Non-A team
© INRIA / Photo H. Raguet
TurtleBot and WifiBot, Non-A team - © INRIA / Photo H. Raguet
The Non-A team develops new finite-time algorithms for robotics and signal processing
Numerous parameters (such as time delay in a communication and frequencies in a periodic signal) or internal variables (such as the location and velocity of a moving object, or derived from a signal) can not always be measured. They are most often estimated by methods which are the asymptotic statistics. The Non-A team has developed an innovative approach to build estimation algorithms that converge quickly and after a finite time, which represents a significant advantage for real-time applications
Mathematical methods associated lead to conventional digital tools that reduce the complexity of the algorithms and make them particularly simple to implement. These algorithms are often more secure and faster than existing algorithms. They apply to many areas, including control systems, robotics, signal processing.
© Inria / E-Motion
The team develops models and algorithmic methods to develop artificial systems such as robots, abilities of perception, decision and action . His work applies to intelligent transport systems, from cars to mobility solutions for people with reduced mobility.
E-Motion offers two demonstrations:
- A robotic wheelchair which navigation is directed by user intent
- Perception embedded in a vehicle: detection and tracking of moving objects.