Presentations of 3 robotics builders and developers. nice conversation with some informative links.
iCUBE is first project introduced by Giorgio Metta .
1 robot 250000 euro, built in 6 months and the robot is open source, so you can do it (if you have the financial means)
Balancing, recognizing objects, can point to the objects when asked, manipulation (using the object) which depends on feedback.
Challenges: materials (very fragile), AI, energy, cloud,
Manfed Hild: neurorobotics reseach laboratory (NRL)
little 15 EUR robotset + 2 to 3 hours you can make a walking spider robot. This allows students to get an idea of a robot.
Step by step robot learning is explained by Manfred: first simulations limited to two or a limited amount of parameters, after that e-robot simulation is stated (all the circuits within the network).
Cognitive robotics: with a really difficult platform (dense one), you can only allow a couple of students on the platform, but if you use a less expensive platform, it allows you to add more students to the platform.
Question: which robot would you buy: mindstorms from lego is really good, but as an entrance to robots any robot (cheap) will do to allow learning the necessary elements that come along with the idea of robots.
Seed Robotics http://www.seedrobotics.com/
Robotics in education: right now only in universities students get opportunities to work with robots. Some initiatives within high school, but there is a big gap related to robotics in education.
Robotics is a very differentiated field: maths, design, physics… which gives it diverse practical fields of study. There can be multiple real world tasks that can be investigated across fields.
But it does (mostly) use a top down approach
With using the visual programming language Python it is possible to visualy build a pogram you have in mind, and then get a coded result.