Servo - March 2014

26 SERVO 03.2014

FORCE(D) TO WORK TOGETHER

We've seen all kinds of robots that have been programmed to collaborate with humans, but few so directly as this Georgia Tech robot — thanks to a wearable sensor that lets it "spy" on its human partner.

“It turns into a constant tug of war between the person and the robot,” explains Billy Gallagher, a recent Georgia Tech Ph.D. graduate in robotics who led the project. “Both react to each other’s forces when working together. The problem is that a person’s muscle stiffness is never constant, and a robot doesn’t always know how to correctly react.” For example, as human operators shift the lever forward or backward, the robot recognizes the command and moves appropriately. When they want to stop the movement and hold the lever in place, people tend to stiffen and contract muscles on both sides of their arms. This creates a high level of co-contraction.

“The robot becomes confused. It doesn’t know whether the force is purely another command that should be amplified, or ‘bounced’ force due to muscle co-contraction,” said Jun Ueda, Gallagher’s advisor and a professor in the Woodruff School of Mechanical Engineering. “The robot reacts regardless.” Most other sorts of robotic collaboration involves passive sensors that are designed primarily to keep the robot from accidentally maiming or killing any humans that might come near it. Actually being able to figure out what a human intends to do as the human intends to do it has the potential to make collaboration both safer and more effective.

The downside here is that you have to outfit humans with a bunch of complex sensors, but that's starting to get to be much less of a big deal. For example, with something like Thalmic Labs' Myo armband, you might just be able to slip on an accessory with a minimum amount of hassle that accomplishes what Georgia Tech's sensors can.

Read more at www.news.gatech.edu/hg/item/266851.

BEST WAY TO MAKE A COOKIE CRUMBLE

Ben Krasnow is on a mission. He’s looking for the perfect chocolate chip cookie. To aid him in this noble endeavor, he’s created the cookie perfection machine. From cleaning with plasma to a DIY CT scanner, to ruby lasers to LED contact lenses, Ben has to be one of the most prolific and versatile hackers out there today. What better way to relax after a hard day of hacking than to enjoy a glass of milk and a perfect chocolate chip cookie? This is actually an update to the machine that first debuted in 2012. Ben has loaded his machine up with ingredients, and has everything under computer control.

The machine will now dispense the exact amount of ingredients specified by the computer, measured by a scale. Everything happens one cookie at a time. The only downside is that the machine doesn’t have a mixer yet. Ben has to mix a single cookie’s worth of dough for every data point.

His experiments have returned some surprising results. Too little flour actually results in a crisper cookie, as wetter dough spreads out to a thinner layer. Ben also found that adding extra brown sugar doesn’t result in a more chewy cookie. Even though he’s still in the early experimentation phases, Ben mentions that since it’s hard to make a bad chocolate cookie, even his failures taste pretty good.