ST5 : Semi autonomous drone navigation Introduction J er emy Fix - PowerPoint PPT Presentation

ST5 : Semi autonomous drone navigation Introduction J´ er´ emy Fix jeremy.fix@centralesupelec.fr Herv´ e Frezza-Buet herve.frezza-buet@centralesupelec.fr Thursday, 1 st september 2020 1 / 25

General information ? ? 2 / 25

General information ? ? 3 / 25

General information ? ? Where to find informations ? 4 / 25

General information ? ? Where to find informations ? • on the website http://st5drone.metz.centralesupelec.fr 5 / 25

General information ? ? Where to find informations ? • on the website http://st5drone.metz.centralesupelec.fr • on the edunao page : expected soon… 6 / 25

General information ? ? Where to find informations ? • on the website http://st5drone.metz.centralesupelec.fr • on the edunao page : expected soon… Ambitious but it is worth it 7 / 25

General information ? ? Where to find informations ? • on the website http://st5drone.metz.centralesupelec.fr • on the edunao page : expected soon… Ambitious but it is worth it • free/open-source sofwares, real robots, and new scientific concepts (designing control loops, handling and updating probabilistic models in robotics, …) 8 / 25

General information ? ? Where to find informations ? • on the website http://st5drone.metz.centralesupelec.fr • on the edunao page : expected soon… Ambitious but it is worth it • free/open-source sofwares, real robots, and new scientific concepts (designing control loops, handling and updating probabilistic models in robotics, …) • but it opens you new perspectives, and you will work with REAL robots in real environments ! 9 / 25

Teaching team I : Come and see us ‼ Shared lectures: Automation and control Jean Luc Collete, jean-luc.collete@centralesupelec.fr Office A002 10 / 25

Teaching team I : Come and see us ‼ Shared lectures: Automation and control Jean Luc Collete, jean-luc.collete@centralesupelec.fr Office A002 Elective lectures: System modeling Virginie Galtier, virginie.galtier@centralesupelec.fr Office B308 11 / 25

Teaching team II : Come and see us ‼ Specific lecture: Autonomous robotics Francis Colas (INRIA Researcher) francis.colas@inria.fr 12 / 25

Teaching team II : Come and see us ‼ Specific lecture: Autonomous robotics Francis Colas (INRIA Researcher) francis.colas@inria.fr Enseignement d’int´ egration Jeremy Fix jeremy.fix@centralesupelec.fr Office B302 Herv´ e Frezza-Buet herve.frezza-buet@centralesupelec.fr Office B301 13 / 25

Major topic : Autonomous robotics Robot : ”An automatic device capable of handling objects or performing operations according to a fixed, modifiable or adaptable program”, Larousse Industrial robots in High speed controllers in deterministic environments partially/fully observed environments . . . Low consumption, Stochastic environnements imperfect and slow devices . . 14 / 25

Machine Learning Machine learning for perception : Deep learning Machine learning for action sequence : Reinforcement learning Video To learn in depth machine learning, come in the 3rd year MDS/SDI major proposed in Metz ! 15 / 25

Seminars Introductory seminars 09/09 PM : ?N. Texier? (ANAFI Project leader) 18/09 PM : G. Beneux (SAFRAN engineer) Simulation and control/command 21/09 PM : J. Soulier (SAFRAN engineer) Drone system : context and feedback on the Patroller 23/09 AM : J.L. Jacquot (CEA Tech Grand Est director) 23/09 AM : A.S. Didelot (President, co-founder) Entrepreneurship in the UAV market: example of a startup All the 23/09 PM : Round table + drink! partners Evaluation 16 / 25

Open source sofwares : Free open source sofwares • Ubuntu operating system ( http://www.ubuntu.com ) • Robot Operating System ( http://wiki.ros.org ) • Gazebo robotic simulator • Git, Latex, Makefiles, …. The machines in the labs are setup but install your laptops as well ! 17 / 25

Open source sofwares : Free open source sofwares • Ubuntu operating system ( http://www.ubuntu.com ) • Robot Operating System ( http://wiki.ros.org ) • Gazebo robotic simulator • Git, Latex, Makefiles, …. The machines in the labs are setup but install your laptops as well ! Ressources • Advised configuration on the st5drone page • Lab works on Tuesday 3 rd • Tutorials : tutos.metz.centralesupelec.fr 18 / 25

Lecture : Autonomous Robotics Content • 10 lectures (15 HPE), 2 tutorials (3 HPE), 7 labs (21 HPE), 1 exam (1.5 HPE). Total 40.5 HPE but 60 HEE. 19 / 25

Lecture : Autonomous Robotics Content • 10 lectures (15 HPE), 2 tutorials (3 HPE), 7 labs (21 HPE), 1 exam (1.5 HPE). Total 40.5 HPE but 60 HEE. Ressources • Lab works and tutorials on https://colas.gitlabpages.inria.fr/st5 robotique autonome/ • Syllabus, lab works and tutorials on edunao + forums 20 / 25

Lecture : Autonomous Robotics Tutorials and Lab Works • Intensive : → SLAM and navigation • Prepare the ROS tutorials before the first practical ! (Friday 11th, see Lab Work 1) • In general, preparatory lab works to be done on your own • (Advanced) Python tutorial on tutos.metz.centralesupelec.fr 21 / 25

Lecture : Autonomous Robotics Tutorials and Lab Works • Intensive : → SLAM and navigation • Prepare the ROS tutorials before the first practical ! (Friday 11th, see Lab Work 1) • In general, preparatory lab works to be done on your own • (Advanced) Python tutorial on tutos.metz.centralesupelec.fr Remote computers In the practicals you will use high performance computers (Kyle) and get connected to it with NoMachine. → Demo More on tutos.metz.centralesupelec.fr soon 22 / 25

Enseignement d’int´ egration General information • Topic : Semi-autonomous navigation of drones • Partner : Parrot • Subject on http://std5drone.metz.centralesupelec.fr 23 / 25

Enseignement d’int´ egration General information • Topic : Semi-autonomous navigation of drones • Partner : Parrot • Subject on http://std5drone.metz.centralesupelec.fr Ressources • Real drones in real environments • Hardwares : bebop2 lent by Parrot • Sofware in ROS, OpenCV • See the video demonstration on the website ! 24 / 25

Enseignement d’int´ egration In practice • Work in groups of 5 students → 6 groups • The subjects are prety well detailed and divided in subtasks • Each subtask is calibrated in duration • Work collaboratively (Git) • Evaluation : oral presentation with a result report (Latex, start it early !) • Parrot will be atending the defense • Your grade will include a group and individual component 25 / 25