Poop scooping robot demonstrates how to make expensive tech really attractive

If you have a dog and an uncrossable back yard, a poop-scooping robot might be worth $400K

Despite rapid advances in artificial intelligence and robotics during the past few decades, and the increasing cost-effectiveness of robotic machines to automate everything from trimming the grass on golf courses to fulfilling orders at distribution centers, robots haven't made much of a direct impact on consumers except for a few robot vacuums and toys.

That may change following the announcement from a leader in the market for general-purpose robots and robot components that it has taught its Personal Robot 2 model how to scoop poop.

The $400,000Willow Garage Personal Robot 2 aren't cost-effective for the average consumer. They're actually designed more as a platform for research into getting robots to take over personal tasks, which involves more programming than most consumers can manage just to give up having to fold laundry or make sausages.

For their pet-management-automation breakthrough, researchers customized the PR2 by adding an extension suitable for scooping the poop of domestic animal companions may change that, however, by taking over some of the more odious aspects of pet ownership.

It was built and programmed by a research team at the University of Pennsylvania's General Robotics, Automation, Sensing and Perception (GRASP) Labs.

The team added a scooper to the PR2's articulated arm and wrote software that allow the PR2 to perceive, navigate and manipulate objects in its environment well enough that it achieved an accuracy rate of 95 percent when tasked in lab tests with picking up artificial dog poop. The research team did not explain why it has artificial dogs in its lab or why it allows them to poop there.

Unfortunately this obvious breakthrough in human/robot relations is weighted down with an overly academic extended acronym – Perception Of Offensive Products and Sensorized Control Of Object Pickup (POOP SCOOP) – a potentially threatening nickname (Graspy) and no current prospects of becoming a commercial product any time soon.

IEEE Spectrum has a nice story on the mod, including things probably best left unimagined – such as the tidbit about the robot being able to identify poop based on color and haptic (touch) feedback to know when it has picked it up.

Potential robot owners who know all too well how to identify Offensive Products by both sight and touch through a doggy-litter bag will recognize the validity of those criteria, but don't care to hear any more about them.

Even at its best, Graspy wouldn't follow pet owners on a morning walk. It's designed to "clear poop out of an open field" according to GRASP Lab team member and Ph.D. candidate Benjamin Cohen, speaking in a video subtitled "20poops in 12minutes."

Though it won't be a product soon, the GRASP Labs team plans to post the software and documentation online at ROS.org so other PR2 owners or researchers can replicate the achievement – preferably before the back yard becomes completely uncrossable.

Read more of Kevin Fogarty's CoreIT blog and follow the latest IT news at ITworld. Follow Kevin on Twitter at @KevinFogarty. For the latest IT news, analysis and how-tos, follow ITworld on Twitter and Facebook.

Top 10 Hot Internet of Things Startups
Join the discussion
Be the first to comment on this article. Our Commenting Policies