Sarah Mellish of Yaskawa Motoman Robotics explains the role of artificial intelligence for robotic order fulfillment:
“Technical innovations in artificial intelligence and robotics have unlocked massive potential for retailers to automate the handling of diverse Stock Keeping Units (SKUs) in the order fulfillment process with human-like flexibility at high speed….”Read More >>
A Neocortex® G2R (Goods to Robot) Cell is easy to deploy, like a collaborative robot. Yet it is uses an industrial Yaskawa Motoman robot with up to twice the capacity and 150% the speed of a collaborative robot (up to 800 items per hour).
How easy to deploy is the Neocortex G2R Cell? The average time for a Cell to be up and running at a customer site is one day. See our fun time-lapse video here. After it is up and running, our team then optimizes the robot and trains your operators. This contrasts with integrating an industrial robot system that may take weeks before being ready to run. The Cell’s deployment efficiency is comparable a collaborative robot’s average setup time, which is about half a day.Read More >>
Robotic material handling occurs in unstructured environments. In addition, the objects to be moved or manipulated by the robot often have never been seen before, vary constantly, and are in random locations. In the past, there was a tradeoff between throughput and flexibility.
Enjoy this short video blog as I explain on the whiteboard the current status of robotic solutions in the depalletizing application workspace, using throughput and flexibility as the main variables. It represents state of the art, industrial grade solutions for vision-guided robots, advanced vision-guided robots, and machine learning-guided robots.
Enjoy!Read More >>
In a recent article, Boston Consulting Group (BCG) points out that spending on robots worldwide is expected to grow from $15 billion in 2010 to $67 billion in 2025. The $52 billion increase in 15 years is a compounded annual growth rate of 10%. They attribute this growth to a convergence of falling hardware prices, performance improvements, and easier application software combined with increased flexibility and finesse. This results in robots being useful in a much broader set of applications than you might traditionally think of – such as automotive assembly and welding.Read More >>
Note from Bob Ferrari’s Post on Aug 7, 2014 entitled Permanent Shifts in Consumer Shopping Trends Have Supply Chain Implications. He comments on a quote in the article: Online Customer Fulfillment, Retail Supply Chain, Supply Chain Business Process that cites the following from Shopper-Trak: “Online sales have grown more than 15% every quarter for the past two years and are having a big impact on the way many companies are looking at their brick-and-mortar stores…. Rather than networks of distribution centers and fleets supporting individual physical stores, the new emphasis will be on high-volume online fulfillment supported by combinations of fulfillment centers and multi-purpose retail outlets.”Read More >>
Check out this interesting article on advances in deep learning for robots: Robots Helped Inspire Deep Learning and Might Become Its Killer App?
It’s worth comparing/contrasting between deep learning techniques and Neocortex. Neocortex is Universal Robotics’ patented machine learning platform, based on a seven-year development effort between NASA and Vanderbilt University. Even though the technology is currently employed on Robonaut 2 on the International Space Station, Universal’s focus with Neocortex is material handling tasks. By learning to recognize new objects or recognize previously seen objects that have changed, Neocortex machine intelligence brings flexibility to material handling automation.Read More >>
Check out this video showing Universal Robotics Neocortex – a patented next-generation machine intelligence software platform. It is guiding a robot to pick up a wide range of parts, providing flexible automation.
Traditionally, flexibility consisted of manually reconfiguring mechanical systems and sensors, and manually re-engineering algorithms to accommodate new parts. Neocortex’s machine learning can automatically handle a wide range of changing parts, reducing the need for manual changeovers. In the past, when a new part was introduced, even if the robot could pick up the new part, the robotic work cell would still require changes in fixturing, sensors, machine vision algorithms, and machine control.
In this video, see Neocortex guide the robot to handle various parts. Also shown is a demonstration of the simple training method for those occasional times an operator needs to teach Neocortex something new – a process that takes less than two minutes.Read More >>
We’ve recently discussed how the Microsoft Kinect can be used in robotic mobile manipulation. See above a video showing what is, to our knowledge, the first commercial integration of the Kinect with an industrial robot. Signal from standard webcams is also used. A software from Universal Robotics crunches the data to obtain a 3D representation of the scene. The application shown is the palletizing of randomly-placed boxes with Motoman robots.
Seeing consumer priced sensor entering into the industrial arena is very exciting, as long as the integration can prove to be robust enough.Read More >>