Humanoid Robot HR-2 239
Denix writes "The HR-2 humanoid robot was constructed during a period of three months at Chalmers University in Sweden.
It has 22 degrees of freedom which enables it to easily move around imitating human motions. The robot is also equipped with stereovision giving it possibilities to perform hand-eye coordination. For that task an artificial neural network is evolved. Furthermore, the artificial brain is capable of tracking faces as well as recognising them. The HR-2 is also able to speak.
The website also contains a movie (35.5 MB) of the HR-2 in action."
Humanoid vs task-specific robots (Score:5, Interesting)
Aren't in most situations robots designed specifically for one task (or a small group of tasks) better?
My Roomba robot vacuum broke a few weeks ago and it took 2-3 weeks to be replaced. If that same robot also washed my clothes, did my dishes and cooked my food, I'd have been in bad shape.
[end serious post]
[start joke]
Come to think of it though, I'd definetly spend serious money on a robot that vacuums the floor, washes clothes, does the dishes and cooks... and does not demand attention, new clothes, a wedding... Heck, let's go make these humanoid robots already! And, of course, they'll need some specific, er... anatomic, parts and capabilites. And they absolutely must have an off and mute switches.
Quantum Consciousness, Not Size, Counts (Score:4, Interesting)
If the Swedes can integrate the microtubes into the neural network controlling this robot, then the Swedes will achieve a sentient cybernetic device. Such an achievement would qualify for a Nobel Prize in physics.
"Impressive. Most Impressive." utters a human-machine hybrid in Star War V.
This is a little more advanced (Score:5, Interesting)
Re:Awesome! (Score:3, Interesting)
Robots are often clumsy, and a big reason for this is inadequate sensory ability. Robots are already good at responding very precisely with their limbs, because they have a computer model of how their limbs react and can sense how they are presently positioned, at least with respect to the robot itself. But clumsiness becomes a factor when limbs must be positioned with respect to outside objects - this requires recognition.
The big deal is that processor technology is getting to the point where a robot can analyze high bandwidth sources like video in real-time, for use as feedback for its limbs, or whatever else it may control. With stereoscopic vision the robot could constantly refactor it's movement based on analyses of video, so it could say, go into the garage and roll out the garbage while avoiding obstacles. Or how about the watering plants? Or returning scattered items to their pre-defined places? Wiping down the kitchen counter? Doing laundry? Robots will play a big role in our future, for better or worse. The market will find them hard to resist once they actually become useful.
Fast digital signal processors will play a role in this, not unlike the 7 SPEs found in the Sony/IBM CELL processor. Fast, commoditized (read: cheap) DSPs for this sort of application haven't been available, and that will soon be changing. I suspect PS3 games will use those vector processing units for responding to and learning from high bandwidth inputs like audio, video, or even EEG data to sense states of mind. I mean they called the thing the CELL after all. Likely because of it's seemingly organic response to stimuli.
Re:Humanoid vs task-specific robots (Score:3, Interesting)
Now that would be worth money.
Dynamic equilibrium (Score:3, Interesting)
This robot and almost everyone I've seen manages the walking by substituting a stable state with another stable state (static equilibrium [wikipedia.org])...
tom