The Science of Folding Clothes

Getting your clothes to fit neatly inside a suitcase can sometimes be struggle, but robotics engineers at UC Berkeley can help you out.

They’ve come up with an efficient way to fold a variety of clothes into neat little rectangles. These techniques are intended to help a new generation of robots take on a monotonous household chore: folding laundry.

Using cameras and shape recognition software, the robot is able to assess the best way of folding each piece of clothing based on the shape of it.

The key to improving robots is in Artificial Intelligence (AI). Robots can only do what their programming tells them to do — and often can’t adapt to new or unique circumstances.

For example, you can program a robot to fold a shirt, but if you throw in a shirt with buttons on the opposite side, the robot may not be able to adapt to the new situation and fold it.

The UC Berkeley engineers are trying to develop robots that don’t rely on such specific programming.

Learn more about these techniques for folding clothes

Making Huge Strides for Mobility

This exoskeleton, developed by UC Berkeley professor Homayoon Kazerooni and his team, helps people suffering from spinal cord injuries to walk again.

“Many paraplegics are not in a situation to afford a $100,000 device, and insurance companies don’t pay for these devices,” Kazerooni said. “Our job as engineers is to make something people can use.”

To make his exoskeleton affordable, he used the simplest possible technology: a computer and batteries in a backpack, actuators at the hips, and a pair of crutches with buttons that activate an exoskeleton that fits around the legs. The crutches provide stability, an important consideration for paraplegics navigating streets and sidewalks.

“The key is independence for these people,” he said. “I want them to get up in the morning and go to work, go to the bathroom, stand at a bar and have a beer.”

Read more

The robot spider that shrieks as it walks

While this eight legged creature is still a prototype, UC Santa Barbara alum Matthew Garten hopes to debut the finished robot for this year’s Bay Area Maker Faire.  Currently the wooden joints in the legs let out a loud squeal that he’s hoping won’t be in the final version (but definitely give off a creepy vibe in his test video).

The technology he’s using is known as the Klann linkage and essentially was developed in the mid-1990s to replace a robot’s wheels by simulating an animal’s walk.

Matthew, seeing himself as both an engineer and inventor, says that robots roam his home.  He’s worked on a wide range of projects from MEMS stem cell sorting to rocket-propelled grenade defense. Also you may have seen Matthew’s open source steampunk Arduino watch on Instructables a few years ago.

Like this post? Follow UC Research on Tumblr.

Building robots to land on Saturn’s moons

Building robots to land on Saturn’s moons

Landing an unmanned robot on another planet can be quite a feat and can end up being quite a complex process.  Scientists want to make this process easier but also allow us to explore worlds that are currently too difficult to land on.

UC Berkeley professor Alice Agogino is working with doctoral students to build what are known as tensegrity robots.  Essentially, these are robots built with a series of rods and tension wires that protect the delicate scientific instruments in the middle.

The structure allows for both flexibility and strength while navigating a rugged environment — for example, landing on a planet’s rocky surface. These robots can explore places that are currently inaccessible to wheeled rovers such as rocky cliffs, which are rich in geological data due to the exposed rock.

Currently, NASA researchers are working on a prototype to one day land on places such as Titan – one of Saturn’s moons.  Scientists are interested in this moon because it has a thick atmosphere with flowing liquids on the surface and is often referred to being the most earthlike world in our solar system.

Read more about this technology 

A whole new meaning to “Deadhead”

Picture 2013-09-17 at 3.33.11 PM
Above is a representation of former Grateful Dead drummer Mickey Hart’s brain.  It will be used on his upcoming tour as both a visual and as a way for his mind to power the stage’s lighting.  Hart explains:

“I was just looking at it and watching it fire, and you see the colors moving and the different rhythm patterns and realizing, that’s me!”

We specify “representation” because the visuals are part of a collaboration between Hart and UCSF researcher Adam Gazzaley.  Gazzaley is quick to say that the images are stylized for the sake of special effects, but the method in which to read Hart’s brain in real time has much larger scientific and medical implications.

Read more here

Robotic squirrels vs. rattlesnakes

Picture 2013-01-24 at 3.42.46 PMUC Davis researchers go into rattlesnake country to study the interaction between snakes and squirrels. The snakes are real, but the squirrel is a robot.

“By bringing engineers and biologists together, we’re creating new ways of doing science.  Hopefully making discoveries that would not have been made unless we brought these different fields together.” – Sanjay Joshi

A number of former and current students have built the robosquirrels over the years, primarily from the S. Joshi and D. Owings labs at UC Davis, in close collaboration with the R. Clark lab at SDSU. Former PhD student Aaron Rundus built the first robosquirrel for his laboratory studies, assisted by Nick Giannini and Erin Chin. Former MS student Ryan Johnson built the first outdoor robosquirrel for field studies, with input from Matthew Barbour. Alex Barszap, Zac Dillow, Armen Davtyan, Laine Tennyson, and Travis Kupp have modified the squirrels in various ways for continuing field studies, with help from Bree Putman.