This is CHARLI, the United State’s first ever human-sized autonomous robot. In contrast to other humanoid robots such as ASIMO, CHARLI seems rather obsolete. However, they have a goal, they have a prototype, and they have a good project leader to take them through the long winding path towards the construction of a robot that can help people around their homes and around the city.
As Awesom-o writes in the Artificial Intelligence and Robotics blog:
Minus the face, it looks quite like the robot from the 2004 movie I, Robot (and story by Isaac Asimov), and that’s because it is partly based on it. The inspiration behind CHARLI is quite simple—we live in a world tailored for humans, and so it makes sense that our robot helpers will look and be able to do most of the things we can. At five-feet tall, it is a little bit shorter than the average human, but CHARLI will be able to walk, run, jump, open doors or squeeze through tight places. Basically, it will be able to mimic us in almost all ways.
To automate human intelligence, it is better to start by building a complete human-like system with the abilities of a human baby (or even an insect!), and progress from there, rather than concentrate on “adult” versions of isolated skills, and then hope that we will be able to eventually glue the various components together. The hope is that a robot able to interact with its environment in all the complex ways that a human can will be able to learn the more advanced skills, rather than have them pre-programmed into it as symbolic reasoning programs.
The interesting thing is that this book was published in 1998, twelve years ago! It looks like the concepts imagined back then are just being put into practice. So little have we advanced? I mean, I concur with the fact that computing speed was not the same 12 years ago and it would have been impossible to do the amount of calculations we can do nowadays. However, robotics as a whole seems to have progressed very little since then.
It seems we still have a lot to learn, and a lot to discover.
One trend I have seen popping up a lot is the fact that all development of humanoid robots tends to be oriented towards “helping people in the house or around the street”. I think that sole aim of building them is limiting research as a whole. I believe it is better to tackle the whole solution to the problem, than to focus on a specific goal. This may be completely debatable, but in my own opinion single-mindedness limits the ability to go further than the initial solution of the problem. Never set limits.
Read more about CHARLI at the Virginia Tech blog.