http://hackaday.com/2010/10/11/learning ... peds-walk/
Thought this might be of use to anybody designing bipedal robots or animating walking robots
Mechanics of Bipedal robots
Moderators: MR.D, Moderators
Re: Mechanics of Bipedal robots
TBH, it's a lot more useful to look at things like Disney tweens.
This isn't a particularly good example of mock-human motion; the robot doesn't use its arms to shift where the center of gravity is or introduce torque, for example, which looks very false and un-natural, and the robot's waist is largely un-involved in the behavior. The problem with looking at anything from today's robotics labs is that they fall very short of the ideal and while it may be fun to have something that moves crudely (ala The Can from DoW) it's a matter of art, not design (The Can makes no engineering sense at all).
Walking is a full-body motion. It's something that a lot of people have trouble with, largely because they walk without thinking. Try walking with your arms stiff at your sides, like that robot, and you'll immediately feel why it's wrong.
For bipeds, you really want to study the most successful models, which are all creations of Nature- us, birds, etc., and animation studies of characters done by the best people.
Best thing I've found for everyday tweaking is to get up and do a motion that the character's doing, and get a feel for where weight would be and how it would effect things. Your Significant Other will have trouble understanding why you want to "understand what those guns will do to my stride", but it can really help you get a better feel for the animation process.
In the end, though, your objective is to do good art, bringing your character to life and giving it a fluid motion. That may mean throwing engineering considerations out the window, but if you aren't aware of the rules, it's hard to know when you can break them.
This isn't a particularly good example of mock-human motion; the robot doesn't use its arms to shift where the center of gravity is or introduce torque, for example, which looks very false and un-natural, and the robot's waist is largely un-involved in the behavior. The problem with looking at anything from today's robotics labs is that they fall very short of the ideal and while it may be fun to have something that moves crudely (ala The Can from DoW) it's a matter of art, not design (The Can makes no engineering sense at all).
Walking is a full-body motion. It's something that a lot of people have trouble with, largely because they walk without thinking. Try walking with your arms stiff at your sides, like that robot, and you'll immediately feel why it's wrong.
For bipeds, you really want to study the most successful models, which are all creations of Nature- us, birds, etc., and animation studies of characters done by the best people.
Best thing I've found for everyday tweaking is to get up and do a motion that the character's doing, and get a feel for where weight would be and how it would effect things. Your Significant Other will have trouble understanding why you want to "understand what those guns will do to my stride", but it can really help you get a better feel for the animation process.
In the end, though, your objective is to do good art, bringing your character to life and giving it a fluid motion. That may mean throwing engineering considerations out the window, but if you aren't aware of the rules, it's hard to know when you can break them.
Re: Mechanics of Bipedal robots
This reminds me...my uni is getting a ShapeWrap III Plus system in the next couple of months and I will be trained in its use by the manufacturer. Might throw some animations up here if I have time.