Are there any robot platforms such as the Sony robot or Honda Asimo that can be purchased for programming & experimentation. These would have the flexability to have programs added to them & some ability to learn & adapt.
I have searched around and did not find much. Also I want something a little more advanced than the 1 foot toy robots that are now available.
Unless your name is Bill Gates you cannot afford either of these robots. They are R&D only, with the Asimo vosting millions of dollars. The Qrios are probably $50,000+ each, if Sony sold them.
Like Joe says, the new breed of bipedal robot kits from Robotis or the Robo-One are anything but toys. They are in essense small versions of the Honda and Sony robots, with the same control problems to solve.
Learning and adaptation are issues for whatever controller you are using. The Robotis and other robots could be refitted as you chose, or even controlled remotely by the Cray in your garage.
Well, the situation is that if I can program a robot to do at least 1 house hold chore then I could present this to venture capitalists to obtain further financing. So it would be better to use a robot around 4' or 5' high to do this.
But there are other things financiers will think about -- like how much business liability is involved in letting a 4-5 foot robot around children and animals. This is why most commercial robots are not used in the home.
If you're going for size, the absolute first thing to program a robot to do is not kill, crush, bruise, cut, or maim family members.
You'll notice iRobot's vacuum cleaner is so light it doesn't pose a problem even for Junior's hampster.
I absolutely agree with you, but how does one explain the robotic lawnmowers? I'd always assumed commercial robots would have to be harmless around a baby, but those mowers look dangerous. Do they make you sign a release to buy one?
You raise a good point, though the fact that they are for outdoor use only would tend to lessen the liability aspects. Still, the (few) manufacturers who build these run the risk of liability from a homeowner who goes in for a beer just at the time when the neighbor child gets his fingers caught under the thing. They all have safety features, but with age and neglect the safety features lose their effectiveness. The homeowner may be to blame, but it's the manufacturer who gets sued.
Again, viewing this as a financier, product libaility could totally destroy the investment, so the risks just may not be worth it. Say you make a $10 M profit. Just ONE accident, even if the suit is not successful, could wipe that out in lawyer fees and bad publicity. Buckminster Fuller had first-hand experience with how a single accident can totally destroy a product, an idea, and even a market.
When the MECHANICS of the robot are such that they don't pose such a high liability, then robotics in the house will be a more attractive business proposition. This isn't a programming issue, it's a materials and construction issue. Robots are just too damned heavy and fragile. In Japan they're developing large robots for elder assistance; however, these require very specific environments, and product liability cases aren't quite as common in Japan.
Just do a google search for robotic lawnmowers. There are about 5 or more manufacturers, and they seem to be getting popular, although they cost as much as a good riding mower. They all use buried-wire boundaries and lead-acid batteries.