This Robotic Odd Couple May Soon Run A Warehouse Near You

Author

Lauren Orsini

May 3, 2015

Sometimes two is better than one—and faster and cheaper, too. That was the thinking behind Fetch Robotics’ new robot helper duo.

Meet Fetch and Freight, two models that work together to make e-commerce shipping more efficient by dividing up the work. Fetch is an intelligent humanoid designed to identify and pick up products from a company’s warehouse shelves. Freight looks like a mobile ottoman that can travel up to 4.5 miles per hour to deliver the products from Fetch to shipping.

“There’s a lot of energy spent moving goods around a warehouse and we’d like to use Fetch and Freight to do all that repetitive transportation and have people do what they’re good at,” Fetch Robotics CEO Melonee Wise told me.

Robots From The Past

Fetch and Freight are the first robots to come out of Fetch Robotics, but if you consider Wise’s work history, they’ve been years in the making. Many of the robots’ features recall Wise’s previous work as a roboticist and founder at both Unbounded Robotics and Willow Garage. For example, both Fetch and Freight can dock themselves at their respective charging stations autonomously when they detect their batteries running low, using technology that was pioneered in Willow Garage’s PR2.

Likewise, Fetch looks like a blue version of Unbounded Robotic’s orange UBR-1—too much so to be a coincidence. You can think of UBR-1 as the very first prototype to indicate the need for a robotic duo instead of just one helper. Wise said that while humanoids like UBR-1 and Fetch can pick up items and deliver them elsewhere, they simply aren’t very speedy.

“When you look at a robot that has high complexity like Fetch, you have a really high center of gravity, so that robot can’t drive very fast,” she said. “We wanted to have a much faster robot that could move through the warehouse very quickly.”

Robots For The Future

That’s how Freight entered the equation. It’s actually the more versatile of the two, as it can work with both Fetches and people. Fetch can lift up to 13 pounds, which covers many but not all of the goods you’d find in a typical e-commerce warehouse. When the goods are too heavy for Fetch, Freight can pair with a human to receive and deliver bigger items using an application called Follow Pick. It’s a manifestation of Frederick Taylor’s Scientific Management propelled into the digital age.

It’s cheaper, too. Since Fetch and Freight aren’t monogamous, a company can purchase, for example, 30 Fetches and 60 Freights to scale up the amount of work.

“That’s another reason why we looked at having two robots,” said Wise. “You can drastically reduce the overall cost by having more of the much simpler, lower cost robot.”

The obvious downside of having two robots instead of one is increased technical complexity. The company’s roboticists have had to program tracking into both units so each robot knows where the other one is and when the other one is in position to deliver or receive an item.

Freight also needs to be able to navigate the warehouse without running into anything and to follow a human worker around without that person having to wear some kind of identifying beacon. 

Fetch and Freight will be on display at at ICRA, the IEEE Robotics and Automation Society’s flagship conference, this May. Fetch Robotics has yet to specify a price.

Research and Development versions of the pair will begin shipping to researchers this June, where Wise predicts roboticists will come up with far more than simply warehouse applications for Fetch and Freight.

 “Yes we’ve targeted this one application, but in the long term we see Fetch and Freight as platforms,” she said. “When we talk about developing applications for the robots, we’re talking about expanding their capabilities. One day maybe they will have the capability to get a drink from the fridge or do your laundry, but that’s very far off in the future. If people do develop capabilities like that with the R&D platform, I strongly encourage it.”

This article was written by Lauren Orsini from ReadWrite and was legally licensed through the NewsCred publisher network.


Comment this article

Great ! Thanks for your subscription !

You will soon receive the first Content Loop Newsletter