Amazon.com, Inc. Hosts Robot Contest To Automate Warehouse

26 Mar 2015 | Author: | No comments yet »

Amazon Is Hosting A Robot Contest to Find Its New Employees.

Some 25 teams will compete in Amazon’s upcoming robot throwdown, a competition that will test the outer limits of what a robot can see, grasp and pack into a cardboard box. Some companies view their employees as an asset; Amazon views them as a problem to be disrupted along the path to perfection (and cheap 2-day shipping on paperbacks).Various robots will be participating in the competition where the smartest one will win $25,000, however the one breaking product accidentally will be penalized Amazon.com, Inc. (NASDAQ:AMZN), relying heavily on automated robots to help workers, is hosting an event to spur efficiency among the robots and help increase the productivity for the company, on the move to ensure the customer receives the order on time.In a move certain to raise eyebrows among Amazon’s army of distribution center workers, the company is to hold a contest to find a robot that can pick and pack items more efficiently than humans. While the Web firm already uses robotic technology for some tasks at its distribution centers, the machines are still rather slow and clumsy when it comes to handling more delicate items or awkwardly shaped products.

An upcoming robotics competition run by Amazon’s Kiva Systems subsidiary could move us closer to the age of the fully robotic warehouse, as about 30 teams from academic institutions square off to see who can build the best robot for packing a box. They’re basically a Roomba mated with a forklift — they get underneath shelves and drag them across the floor, reducing the distance employees have to walk. Yes, it’s the X Factor of the robotic world, where precision and care when handling goods will win, rather than any previously unseen singing and dancing ability. Approximately 30 teams will contest the Amazon Picking Challenge which will take place at the International Conference on Robotics and Automation in Seattle.

Points will be awarded for computer vision — the ability to tell apart a box of Oreo cookies, for instance, from a box of Cheez-It crackers — as well as dexterity. The bot-based challenge, which takes place in Seattle in May, will present the participating machines with various tasks to see how effectively they can locate, retrieve, and pack different items. The video above demonstrates the entry from the University of Colorado Boulder, which uses a commercially available Baxter robot armed with several modifications to enable it to carry out the sensitive tasks required. According to the specs, it has to be able to locate products on a shelf, pluck them out of containers, and pack correctly into cardboard shipping boxes.

The robots have certainly helped workers in reducing the workload by not wasting time in finding the products; however, no robot has yet been able to compete with the speed and dependability of a human which makes this contest all the more interesting. If we can teach robots to learn how to cook, anything is possible, but the deep learning techniques needed to make this happen aren’t easy to pull off. Although Amazon hasn’t announced plans to formally incorporate the winner’s robot into its business, there is a good chance that a deal will be struck if the technology proves effective. Robots already play an instrumental role in Amazon’s packaging centers, ferrying 700-pound inventory shelves in and out of storage, but the challenges of handling individual objects with care has posed a persistent challenge for researchers. You can easily design a robot that’s more dexterous than a human and doesn’t require rest (or a paycheck), but that only holds true for extremely repetitive actions.

This could help ease employee workload during the busiest times of the year, or ultimately lead to full warehouse automation and thousands of lost jobs. Amazon says it hopes the contest will “strengthen the ties between the industrial and academic robotic communities and promote shared and open solutions.” Robotics has come on leaps and bounds (quite literally) in the last few years, with advancements in sensor hardware and recognition software making a picking robot actually feasible. Whether any company manages to impress Amazon enough to turn its warehouses into complete Skynet supply dumps remains to be seen. [MIT Technology Review]

Few teams are working on their own robots whereas others are modifying the robots with their own software and grippers that are commercially available. The recent introduction at some of its centers of 15,000 Kiva-made robots, for example, is thought to be saving the e-commerce company between $450 million and $900 million a year.

Aside from replacing factory jobs, developments in self-driving cars could prove massively disruptive to the taxi industry, while the first robot-staffed hotel is also due to open later this year in Japan. Humans have been very efficient with such kind of task specially identifying an object and know how to utilize all their resources to get the task completed. Although robots from Kiva Systems, Amazon’s in-house automation partner, are becoming increasingly productive, they are also becoming increasingly dated and unable to replicate many human actions, like picking and sorting specific items. Wurman said, “We have struggled to pick out a variety of products that represent our catalog and face challenges.” Robot technology has come a long way in the past few years. The winning robots will probably also get some attention from Amazon going forward as it continues looking for ways to automate its fulfillment centers.

Here you can write a commentary on the recording "Amazon.com, Inc. Hosts Robot Contest To Automate Warehouse".

* Required fields
Twitter-news
Our partners
Follow us
Contact us
Our contacts

dima911@gmail.com

ICQ: 423360519

About this site