i wouldn’t let this anywhere near my house either. Perhaps it’ll make a career in hollywood.. as a horror-movie monster

KILL IT WITH FIRE!

The video explains a bit why ‘archie’ is broken.

The head falls of because the robot has ‘jet-lag’ – IT was disassembled for the transportation from canada.
Moreover the robot has serious problems with keeping its balance, because the tilt sensor does not work/was broken during transportation.

The robot has been developed for 4 years now and will be ready in about 5-7 years.
The robot shall (in future) ressemble a common 14-year-old european teenager *lol*

I agree. Why do these researcher put a human face on robots? Most of the most beloved movie robots were not made to look like humans. Give me a R2-D2 any day over the creep fest above.

I think a lot of humanoid robot researcher make the common mistake of confusing visual realism with functional realism. Visual realism runs into serious problems, because humans are very good judges of what looks “human”. There is a well-documented effect where people react far more negatively to a near-photorealistic representation of a human than to a stylized/abstracted representation (cartoony, etc). Humanoid robots need to be able to do what humans do, not look exactly like them.

It will be interesting to see what it looks like when it is finished…! Will it be creepy or uncanny? Hmm!

@Moose: I’m assuming you’re talking about the “uncanny valley” effect. Have you really found it to be “well-documented”? I find it to be very intuitive, in that it reflects what many people experience when they look at these robots. However, the scientific literature I found has actually very little basis (Moris original paper has something like 10 persons in the experiment or so).

The reason I’m asking is that there appears to be very little insight into what is and what isn’t acceptable to humans, and most importantly /why not/. That is why so many robotics research keep making the same mistake (not all, though ;-)

Of course, one could say that: Just don’t make them look like humans at all and you’re set. However, there is a reason roboticists try to emulate humans: They are trying to make robots more interpretable to humans, which is still a big problem today.

R2-D2 and C3PO are actually excellent examples of what to aim for: Understandable, acceptable, non-human but human understandable emotional expression (just consider R2-D2’s chirps and C3PO’s body language!). The problem, of course, is that they were controlled and voiced by humans and nobody knows how to generate that kind of expression automatically, so far.

@Ingo: Ooop! Think you caught me there, that probably wasn’t the best choice of words. Coming from a animation background, I’m used to it being the consensus, and sometimes I’m prone to assumptions. After a little digging I found a few documents regarding it, one from MIT (http://www.mitpressjournals.org/doi/abs/10.1162/pres.16.4.337), and another from ACM SIGGRAPH(http://portal.acm.org/citation.cfm?id=1400921)

I agree that a certain level of anthropomorphism is necessary for effective human-robot interactions, indeed it will be one of the major sticking points in regards to widespread adoption of robotics into everyday life. The question isn’t whether or not we should make robots with humanoid qualities, but to what degree.

This design is in danger of looking way too close to a severed head on a pole. Wonderful.

From the look of that robot, I guess the designers were inspired by the film Saturn 3! (A good film if you’ve never seen it?)
http://www.imdb.com/title/tt0079285/

I always think robots in the future need to look like Bender’s, simplistic face.