CPE Lyon work from mid 2015 to be ready for robocup@home 2016

Robocup@Home experimentation area

Robocup@Home Overview

The robocup@Home challenge is one of the major league of the robocup competition. The aim of the robocup@Home is the realization of set of simple actions into a complexe environment in contrast with RoboCupSoccer league where robots accomplish complex tasks into simple environment.

Robots in robocup@Home evolve into a « home-like » environment. They needs to accomplish tasks like interaction with human into natural language, human following, navigation into dynamic environment, learning where set of objets are,  pick and place actions according human order.

CPE Lyon School into the race

With the help of students , teachers and researchers , Cpe Lyon forms a team to target the robocup@Home challenge. The current design of the robot was designed and is composed of a set of home made parts and commercial products.

Mobile plateforme is a commercial product of PAL Robotic PMB-2 available to support a payload of 50kg
Robot « skeleton » and arm is a home made structure composed of a prismatic elevator, a set of Dynamixel motors and a 3d printed gripper
A RGB-D camera Xtion PRO Live will be used to detect and localize objects
The human localization and interaction will be made with the help of a Kinectv2 RGB-D probe

Cpe robot overview

Research-Teacher in the starting blocs

Students involved into functional parts

Dedicated intership students



Human-Robot Vocal Interaction

The Human and Robot vocal interaction is composed of tree functional subsets: Speach to text, Text processing, Text to speach. The speach to text subset process the voice signal into text according to a given language (dictionary and grammar) and additional... lire plus

Robot Navigation

The navigation challenge has already been addressed by the CPE Lyon team during the robocup@work challenge. The current video shows some navigation processing using the Kuka Youbot robot  and the navigation layer of ROS. Note that in the current configuration, the... lire plus

Object Recognition

Object Recognition and Localisation Concerning the object recognition part, we use the Xtion PRO Live in order to combine 3D point cloud and RGB data to discriminate objects and localize them. The first step of our process is to isolate objets on a plane surface. To... lire plus

Object Pick and Place

Our robot arm is a 7 DoF custom arm made of 1 prismatic joint (elavator) and 6 revolute joints. The prismatic joint is moved by a DC motor and and a ball screw. The 6 revolute joints are made with MX-106 and MX-64 Dynamixel motors. For the moment, the gripper is made... lire plus

People Detection and Following

Tracking people through KinectV2 The people tracking task is made thanks to both KinectV2 skeleton extraction and pedestrian detection through laser information (not yet implemented). We use the KinectV2 to take advantage of its API by getting information like general... lire plus

Sound Localization

Sound Localization trials Concerning speech recognition and audio detection test, the work is in progress. The current test solutions are : Concerning the sound localization, an ad hoc microphone system based on 8 MEMS MP34DT01, 2 shields X-NUCLEO-CCA02M1 and... lire plus

Research Focus

Concerning the research project. Jacques Saraydaryan and Fabrice Jumel are associate members of the Chroma team, (leads by Olivier Simonin) in the CITI INRIA Lab. They work on management of fleet of robots in crowd environment. A general framework called... lire plus

Partager cet article