Industries Logistics News Robotics Technologies Work

Latest ‘Robo-Picker’ Technology Could Replace Warehouse Workers, in Time

Latest ‘Robo-Picker’ Technology Could Replace Warehouse Workers, in Time February 21, 2018 3:35 pm
Latest ‘Robo-Picker’ Technology Could Replace Warehouse Workers, in Time

Photo Credit: Melanie Gonick/MIT

It’s no secret that massive e-retail companies, namely Amazon, are working on systems of warehouse technology that can replace or assist the massive team of humans they currently employ to sort and send items and packages. In fact, Amazon has held annual competitions in which researchers gather to display their latest versions of pick-and-stow technology.

Now, despite the reality that object-manipulating robots have been notoriously difficult to perfect, researchers from Princeton and MIT have edged closer to the precipice of a robotic warehouse takeover. Dubbed the “pick and place” system, a robotic arm outfitted with a suction device and custom gripper has exhibited the capability to pick up and carefully place a range of objects in a manner that outpaces its competitors.

The researchers’ “object-agnostic” grasping algorithm is the underlying code that allows the robot to do what it does, assessing a bin of random objects to determine how to best grip/suction them. What differentiates the “pick and place” system from many of its predecessors, including previous Amazon winners, is that it doesn’t require prior information about or images of an object to effectively assess how to go about manipulating it.

Once the object is in the robot’s grasp, mounted cameras take several pictures of the item, again using an algorithm to match that item into the best-fitting, corresponding bin for storage. A cataloged item list with corresponding, scan-able tags on each item will allow the robot to sort like products in a specific place, despite not being able to differentiate the items by sight alone.

We're comparing things that, for humans, may be very easy to identify as the same, but in reality, as pixels, they could look significantly different, says Alberto Rodriguez, the Walter Henry Gale Career Development Professor in Mechanical Engineering at MIT. We make sure that this algorithm gets it right for these training examples. Then the hope is that we've given it enough training examples that, when we give it a new object, it will also predict the correct label.

It’s true that the technology has applications beyond warehouse sorting, but it’s hard to imagine somebody employing such a machine just to put away their groceries. Clearly, the warehouse is the primary, obvious application for such a sorting system. Still, we never know what the future holds, or how ubiquitous such a technology could become.

This can be applied to warehouse sorting, but also may be used to pick things from your kitchen cabinet or clear debris after an accident. There are many situations where picking technologies could have an impact, Rodriguez added.

Comments