Bridging the robotics divide: Will Durfee

Will Durfee cautions that his inventions pale in comparison to the fictional technology that transforms genius Tony Stark into superhero Iron Man. But Durfee does acknowledge the wide range of possibilities for robots that work in cooperation with humans to make them stronger, and he is developing wearable robots that augment human motion. Although he doesn’t anticipate his inventions will enable humans to punch through concrete walls, he hopes robotics can someday make us all more powerful.

Human powered

Mechanical engineering professor Durfee and his team are leading a project through the seven-university Center for Compact and Efficient Fluid Power to design the first untethered, lightweight, hydraulic-powered ankle orthosis that will power up human motion. The small robot could someday help people with physical impairments walk, or could speed up the typical human’s strut.

"We're looking at what would happen if you're wearing something that is equal to the power of your ankle," says Durfee.

"When we think about wearable robots, we tend to get excited to blast through walls. But humans are not particularly powerful when it comes right down to it.”

During everyday motion, humans put out 100-150 watts of mechanical power — the equivalent of the energy emitted by one incandescent light bulb. Even at peak aerobic performance, the body only reaches about 300 watts. But if a wearable robot could ramp up the body’s power by a meager 100 watts, it could vastly increase human longevity. Durfee envisions that workers engaged in manual labor may be able to perform an activity for twice as long.

Putting robots to work

Autonomous robots that supplant human workers have forever changed the way we make, package and ship goods. But limitations in cognitive processing hinder robots’ abilities to perform complicated tasks. Durfee predicts robots that work in cooperation with humans could bridge the gap and boost productivity, reduce labor injuries, and satisfy the sweet spot for tasks too complicated for a robot.

“In many fields, you need the knowledge and decision-making power of a human, which is difficult to replicate in a machine,” he says. “Picking an apple from a tree involves incredible mind-hand-eye coordination — to use the correct force to pluck the apple without damaging it. Yet, it’s still a backbreaking task. That is where robots can help us.”

Researchers are working to pair the cognitive capabilities of humans with the power of robots. The university’s strengths in mechanical systems, computer science, neurosciences, human motor control, and design will position it as a leader in this promising new field.

Gadgets in store

Durfee’s robotics technology is still in its infancy. His team is doing computer simulations for now, but hope to deliver a tangible device within a year. 

"We engineers can't resist building practical devices,” says Durfee. “So we’re getting underway with doing the beginning design phases of the first real prototype." 

There have been myriad advancements in lightweight actuators, small power supplies, distributed sensing technology, methods of computer simulation and neuroscience that will propel robotics into new and exciting territory. Durfee knows that it will take time for these technologies to mature, but is confident that 2013 will be the year we get friendly with robots.

The Minnesota Discovery, Research and InnoVation Economy (MnDRIVE) initiative seeks to establish a new, ongoing partnership between the U of M and the state. The aim is to discover new knowledge through scientific research to advance Minnesota’s economy, position the state as a leader in key industries, and improve quality of life for all Minnesotans. The university will ask the state legislature to provide $18 million beginning in 2013 to support scientific research in four critical areas, including robotics.

Reprinted with permission from Business @ U of M, a publication of the Office of the Vice President for Research.

Share