Readers be warned, this could end up as a deadly dull essay on a term project i did recently, and which i shall be presenting shortly to my lab. If your sufficiently bored with life to be interested in this kind of thing, please proceed right ahead.
I did this fancy project for my Introduction to Robotics:MEA453 course (Yes, I did introduction to robotics in my penultimate semester of a Masters in Electrical Engineering -with a Robotics specialisation and, no, i really didnt score that well, goes to show how much I really know in robotics). Anyways, it seemed like an out-of-world idea at that time, about 3 months back, but this error has been kindly corrected by my senior who told me such stuff has been in existance for ages now, in the field of humanoid robotics, its just that we dont use it on our pretty little bot.
The basic idea maybe be summarised as thus. Most of our walking can be explained simply by a big swinging ball on a stick. our body when on one foot pretty much behaves like whats given the fancy name of 3D Linear Inverted Pendulum which looks a bit like this:
I did this fancy project for my Introduction to Robotics:MEA453 course (Yes, I did introduction to robotics in my penultimate semester of a Masters in Electrical Engineering -with a Robotics specialisation and, no, i really didnt score that well, goes to show how much I really know in robotics). Anyways, it seemed like an out-of-world idea at that time, about 3 months back, but this error has been kindly corrected by my senior who told me such stuff has been in existance for ages now, in the field of humanoid robotics, its just that we dont use it on our pretty little bot.
The basic idea maybe be summarised as thus. Most of our walking can be explained simply by a big swinging ball on a stick. our body when on one foot pretty much behaves like whats given the fancy name of 3D Linear Inverted Pendulum which looks a bit like this:
We walk by falling. We fall like the shiny golden ball mounted on a stick. Our legs together act like the stick.I must admit, life seems a bit contorted when you think of yourself as a shiny golden ball though. The ball is our center of mass. Playing around with forces and momments acting on this ball pretty much explains everything we see such as Shakira's or Sreesanth's booty shake, the Elvis Pelvis, to my own fluid dance moves (really!). Now the thing with robots is , we like to pretend that this center of mass is stuck inside the pelvis. This makes thinks really simple to calculate, and in no time of approximately 4 years, we can get our l'il robot Hansaram (literally 'one korean guy' - I didnt name it so dont ask) start walking dancing and waving to kids, while we are talking to the pretty ladies nearby. Now I dont know too much about what Hansaram thinks about this deal, but I really dont think thats fair. Everybody loves a good dose of freedom.Least of all, your center of mass is entitled to it. Mine likes to wander about and take a good look around, usually scaring innocent bystanders in the process. The first sign of danger, and the poor thing runs right back into the safe confines of my belly.
Anyhow, what I did was, I took a simulated biped. Did about 9 pages of tedious calculations and tried to find how much the center of mass gallivants about while its poor body was trying to walk. Turns out it wanders out quite a bit without the hapless owner having a say on the matter. The final simulation had big errors mainly i think due to my assumptions on the inverse kinematics (thingie which works out the angles the feet and legs need to make to move about), basically my upper body toppled as i walked. Thats a bug i need to fix, but i still managed to nail the basic relationship. It was tedious but fun.
I am just back from presenting this bit to the lab, telling them that hansaram needs this thingie too. presentation started with a minor glitch with prof getting offended slightly by my hand in pocket stance but things smoothened out and managed to end with applause effectively droning out a seniors pertinent complaints that the results are meaningless, on account of programming errors. Im sure of the math though, so i guess bugs can be fixed. Professor now wants me to think about an implementation for hansaram. I want to go sleep.
If your still hanging around then i guess you must really be perseverant. In which case, I grant you a once in a lifetime opportunity to see my report and play around with my source code. My personal webpage is currently experiencing slight technical difficulties, but im fine with emailing the stuff, so just let me know, and Id be glad to help with additional info.
1 comment:
Allah !!
Post a Comment