Real Steel - How cool would it be to make it a reality!!

 

As a roboticist there is no way you can come away from watching the movie Real Steel (
http://steelgetsreal.com/) without being excited.  I enjoyed the movie a lot, but not for the same reason the other 2.7 million people did (it has topped the box office the last two weekends).  While others were enjoying the movie for its great special effects, or human interest story, I was imaging a whole new hobbyist robot competition.  I had visions of Kinect controlled humanoids shadowing their creators.  Robots about two feet tall fighting in a small ring, being controlled via complex software, a Microsoft Kinect, and a human with a small amount of boxing skills complemented with a huge amount of programming skills.

 

 

 

 

Of-course, this technology already exists (see above video) and remote control humanoid fighting has been around for years.  It's just a matter of combining the two.

 

 

As most of the readers of this blog know, Robo One is a very popular sport in Japan.  For those of you not familiar with Robo-One, it's a fighting competition between two remote control humanoid robots.  The robots range in size from 1 to 2 feet tall.

For more information, check out this article from Robot magazine:  http://www.botmag.com/articles/robo_one_ten.shtml

 

How cool would it be to see some of these Robo-One robots controlled by a Kinect boxing in a ring? How much cooler would it be if the robots were three feet tall?

 

We have the technology, just not the battery capacity.  What if the boxing rules allowed the robots to be tethered?  That would take the power supply burden off the mechanical design of the robots.  The goal would be humanoid robots about three feet tall, with big feet.  The rules would have to regulate the size and weight of the feet.  Heavy enough to allow the robot to both take and give a punch.  Lite enough for the robot to walk (a little) and actually fall down when punched hard enough.

 

The robot would have to be controlled entirely using a Kinect, or a mechanical motion detecting device (body suit).  No controller pads, in fact the human "fighters" would have to wear boxing gloves (trainer size).  Also,  the human "fighters" would have to view the competition from the viewpoint of their robot through a camera mounted in the robots head.  The human fighters would not be able to see the ring directly. This is similar to the mech warfare rules ( http://mech-warfare.com/ ).

 

The audience would be able to see the ring, feeds from both robot POV cameras, and live video of the human "fighters".  How cool would that be to watch!!!

 

Scoring would be done by knockouts, and accelerometers/"hit pads" built into each robot.  The "hit pads" would use the same technology as the pads used in mech warfare, but the data would be supplemented with 3-axis accelerometers to measure the strength of the hit.  

 

Remember, the robots are tethered, and get power from a ringside power supply.  The ring would automatically temporarily reduce the voltage of a robots power supply in proportion to how hard the robot is hit, and where.  The power supply would also decrease slightly every time the robot throws a punch.  This is designed to simulate a fighter getting tired.  It also creates a new strategic component to the competition. 

 

It's times like this that I wish I was not so mechanically declined.  This competition is perfect for the people who have the magic gift mechanical design expertise.  Hopefully the right people read this, and sometime in the near future we can do some robot boxing.

 

 

 

 

 

Interesting research paper on boxing robots:  http://www.cc.gatech.edu/~mstilman/class/RIP08/FINAL_PROJECTS/Misha...

 

 

 

Views: 3261

Tags: boxing, humanoid, kinect, mech, real, robot, steel, warfare

Comment by John Dolecki on October 17, 2011 at 9:25am
Count me In I can do the mechanicals and electrics if someone wants to help with the programming.
Comment by adam pennington on October 19, 2011 at 11:29am
I'd estimate one of the bots from Real Steel would cost around half to one million. The Pr2 cost a quarter mil and is about half the size of what you'd want to boxing bot to be. This was a great comparison between what tech we have going today and in Hollywood. I was thinking along the lines of familiarizing kids with fighting bots as a possible future military strategy. The Talon Robot currently deployed in the middle east is pretty much a remote controlled tank with a machine gun. I would much rather watch robot kill robot than human.
Comment by John Signorino on May 13, 2012 at 6:11am

Briefly, I've always enjoyed using speed bags, double-end bags and heavy bags to train with for self-defense and exercise. What I didn't like is nothing punched back at me. I've been in a lot of street fights in my younger days and am very aware of boxing, MMA, etc, since I study body strikes and movement in real time. I'm also aware of damage to the brain and what strikes to the head are doing, so real speed punches are damaging and sparring against someone to train with is not good exercise for health in reality.

 

Now, I box against a "robot"4 rounds a day I've been developing for the last several years that punches at real human speeds but with about 8500% less force than a human punch. It can punch at superman speeds too. No other device does this. As it is, it is a challenge to box against and can adjust in speed to be very hard to box against. I can even put more movement to it but what for, it's tough as it is.

 

All that I say is true and it does much more. Yet, I bet if I showed you a video of the device you'd be like most and not even respond. Why that is I don't know for sure. I solved a problem by my approach and it works as I say, but people want things a certain way even when it does not matter which way, just a way that works is the important thing.

Comment

You need to be a member of buildsmartrobots to add comments!

EMGRobotics

© 2014   Created by eric gregori.

Badges  |  Report an Issue  |  Terms of Service