Animal Robots

hamlet2600 writes "The New York Time is running an article all about how animal like robots [Soul Sucking registration required] are beginning to become more imporant in furthering research. For years reseachers have been trying to make humanoid robots, Honda's ASIMO, MIT's M2 are some notable ones. It seems that more and more researchers are turning to the animal kingdom for "simpler" means of locomotion."

LEGO Mindstorms are instructive too ...

[YetAnotherName]

My daughter and I came to the same conclusion as the researchers in this article: after struggling to make a biped robot with LEGO Mindstorms [mindstorms.com] robotics for quite some time, we found that a six-legged ant was much simpler.

Why does it have to be snakes?

[rotorhead]

Snake robots have been around for some time www.snakerobots.com/main.htm

Efficiency

[devilsadvoc8]

Bipedal movement is more efficient than quadrapeds. It takes less energy to move the same mass at the same speed using two legs vs four. The problem lies in the inherent instability of bipedal movement. Thankfully, evolution has blessed us with the means to account for this instability. Roboteers don't have the benefit of millions of years and thus an easier solution would be to revert to the less efficient mode of movement involving more than two legs.

Karma whoring!

[martingunnarsson]

Clickable link [snakerobots.com].

Re:sensors and subprocessors.

[Omega697]

problem is that they have YET to design a sensor like our inner-ear to detect balance and orientation. It's called an accelerometer, and not only have they designed it, but they have it in the Sony AIBOs that we use for RoboCup soccer. I'm not saying they're perfect, or even that we know how to use the information from them very effectively. But they do exist.

Re:sensors and subprocessors.

[Lumpy]

no, it is not.

an acelerometer can not detect rotation like the inner ear can nor can it reliably and with extreme precision detect acceleration in 3 axis.

your ear can detect acceleration in very VERY minute detail, along with anlge of tilt as well as rotation.

Granted prolonged rotation confuses the sensor.

I've messed with accelerometers, they are way too low resolution and limited. and the highest resolution and sensitivity units are so expensive that amost all robotics projects do not use them.

The inner ear is more than an accelerometer. plus, it's only one sensor in a group of sensors that animals and humans use for balance.

No Registration Link to article

[krgallagher]

Here [nytimes.com] is a no registration link to the article. This link was generated by New York Times Link Generator [blogspace.com].

ARL at McGill develops similar robots.

[francisew]

The Ambulatory Robotics Lab at McGill develops several robots, including a series based on cockroaches. They work really well... I'm biased, my girlfriend is doing her masters about one (aqua).

I think they have been slashdotted once already... They've got video of the robots online.

If interested, try: http://www.cim.mcgill.ca/~arlweb/Welcome.html [mcgill.ca]

IMHO, these are damned cool!

Re:sensors and subprocessors.

[Animats]

problem is that they have YET to design a sensor like our inner-ear to detect balance and orientation.

Wrong. Buy a small INS here. [xbow.com] There are standard units that contain three accelerometers and three rate gyros (one for each axis), which is what you need. They're getting smaller; 1 cubic inch units with all six sensors are available, and a single-chip version has been prototyped.

Most serious robotics projects today have one of these. They're not good enough for navigation by themselves, but they can provide attitude info just fine.

The basic way you figure out "down" is by using the accelerometers for the long-term component and the rate gyros for short-term corrections. You do lose accurate "down" if you go round and round in a circle for a while. Some aircraft artificial horizons have that problem.

The field has regressed in recent years

[Animats]

Since the MIT Leg Lab tanked, there hasn't been that much interesting work in the US. Insect-level locomotion has been done many times. With six legs, almost any control approach will work. The same goes for swimming robots. True balancing machines are harder. But they've been done.

Raibert did some great work in the Leg Lab's early days. Raibert's big insight was that balance is more important than gait, and he did work with one-legged machines with springy actuators to force the issue. In his day, the Leg Lab had one, two, and four-legged running machines. But he left MIT to do a startup [bdi.com], which seems to have ended his dynamics work. BDI does mostly kinematic models.

The next professor to head the Leg Lab was Gill Pratt, who was more of an actuator guy. He didn't accomplish too much, and is now at some lesser school. Under Pratt, the Leg Lab backed down from running machines to walking machines.

There was somebody after Pratt, but apparently the Leg Lab is now defunct. It's sad. They made so much progress under Raibert.

It's possible to go beyond walking and running on the flat. Legs are really for traction control. All the MIT work assumes that the "feet" don't slip. That doesn't work on real hills or slippery surfaces.

There's two phases to dealing with slip. First, you need to limit joint torques to below where the feet start to slip. Once you do this, you can climb some hills. (Video, 8MB .mov file). [animats.com] That work is ten years old, and still, nobody else seems to be handling leg slip at all.

The next step is to use the three joints of a leg [animats.com] to adjust the vector at which the normal force is applied to keep the ground contact inside the friction cone. Then you can climb more serious hills. Once you get this figured out, much of how humans move when dealing with terrain becomes clear. Leaning forward and bending the knees more when going uphill is all about slip control. Think about it.

Working on this diverted me off into physics engines, because everything that was available ten years ago sucked. So I did a physics engine that worked [animats.com], which turned into a business. There are still very few physics engines good enough for legged locomotion work. Most physics engines, especially the Baraff-type impulse/constraint ones, don't do friction well. Since legged locomotion is all about managing foot-ground friction, you need a simulator that gets friction right. (Hint: if a simulator can't do a driving game without special-casing the wheel/ground contact, it won't work for legged work.)

All this is patented [animats.com], of course.