Robots becoming more independent and autonomous
With swarms of these robots working together but separately, the system can determine which robot is in the best position to retrieve the next bin in an order at any given time. Downtime is virtually eliminated because there’s always another robot nearby if one gets stuck or needs to recharge its batteries. Battery charging is also autonomous.
This autonomy comes from sophisticated fleet management software.
“Think of it as two brains,” said Joe Zoghzoghy, mobile robotics manager for Bastian Solutions LLC, a global material handling and robotic systems integrator based in Indianapolis. “One big brain is controlling all the robots and each robot has a smaller brain that is communicating back and forth. The autonomy is coming from the bigger brain, which is the master control or warehouse control system, and that is coordinating the moves (among each robotic transport vehicle). Then you have smaller brains on the vehicles that are handling the low-level control, for example driving and navigating.”
That’s the swarm effect. The smaller-brained mobile robots are coordinating with each other and with the bigger brain to accomplish certain tasks. The Perfect Pick goods-to-person system by New Jersey-based Opex Corporation exhibits a similar swarm effect. The robotic delivery vehicle called iBOT works in like-minded swarms to deliver bins of goods to manned order fulfillment stations.
Bastian’s director of advanced technology, Derek Cribley, said that with the variety of autonomous robotic carrier systems on the market, each has its strengths and advantages. Criteria range from different transfer rates and payload capacities, to different storage densities. Bastian works with their customers to select the best technologies for their particular solution. He says the latest generation of robotic transport vehicle systems offer even more flexibility and efficiency.
“The Servus system is an Austrian company,” Cribley said. “They already have 15 to 20 systems installed in Europe. We’re now implementing the first two systems going into the U.S. market. The Servus system was the first one that I’ve seen that was able to work inside and outside of an aisle. It has the ability to do not just automated storage retrieval, but also probably replace other technologies like conveying systems or transfer vehicles to bring the product more directly to the work area.”
The heart of the Servus system is the autonomous robotic carrier (ARC), which can leave the aisle and follow curved tracks around the warehouse facility. It can even be outfitted with an onboard robotic arm for item picking while on the fly.
Cribley is especially intrigued by this idea. “It can be performing work while it’s transporting at the same time. You can really reduce your cycle time.”
Bastian is taking that idea several steps further and working on prototypes for combining six-axis industrial robotic arms onto high-payload mobile robots. More on this next-gen solution later.
An intralogistics solution teams humans with industrial robots, enabling direct cooperation between man and machine with no safety fence required. automated item pick (AIP) combines Swisslog’s automated goods-to-person warehouse systems, such as CarryPick and AutoStore, with a robot picking station to assist human workers with picking tasks. Robot picking is performed by a Kuka LBR iiwa robot arm mounted on a mobile platform (flexFellow) that is manually wheeled into position at the picking station.
“We have everything integrated into one station and we can put it into an existing workstation designed for human picking,” said Kirt Laeske, product manager robotic solutions, Swisslog AG. “Just plug this in where a person is standing to help him do his work.”
The process starts with a robotic carrier retrieving a bin and delivering it to the picking station. Each bin has three or four compartments containing different items. The vision guided robot arm finds the right item, picks it from the bin, and places it in a designated area for further processing by the human picker. Robotic bin picking technology has come a long way.
“Typically we are handling from 10 up to 100,000 different SKUs,” Laeske said. “The vision must be intelligent to detect items that have not been taught before. Right now we’re testing a gripper that is able to pick and place a wide range of SKUs.”
He says the system uses a combination of 2-D vision and 3-D scanning. But the real “magic” is in the robot arm. The Kuka LBR iiwa lightweight robot has a kinematically redundant 7-axis arm with exceptional flexibility, designed to work in tight spaces and in close proximity with its human collaborators. The LBR has force torque sensors in all seven joints, making it extremely sensitive.
“It actually feels when it touches something,” Laeske said, explaining how that helps the robot find items in the bin and safely operate side by side with human coworkers. “If there is a failure, the worker can touch the robot, move it by hand, do the correction, touch it again, and the robot will continue working. The LBR iiwa opens up a new field of interface between man and machine.”
“Whether it’s Automated Item Pick, or CarryPick robots, or robotic depalletizing, we are not developing these robot solutions for selling robots,” Laeske said. “We are creating these solutions because we want to make our intralogistics solutions better and more competitive by adding robotics.”
This is not far from the way Amazon made its foray into the robotics world. And they’re not the only big player making strategic robotics acquisitions in this fast-moving logistics space.
Mobility and autonomy
IAM Robotics has developed Swift, a mobile picking robot that combines an autonomous mobile vehicle with a vision guided robot arm in one integrated system.
Established in 2012 and based in Pittsburgh, Pennsylvania, IAM Robotics unites three research engineers from Carnegie Mellon University (CMU). Founder and CEO Tom Galluzzo, hardware lead Vladimir Altman, and chief software architect Ricky Houghton are working to bring mobile picking to life.
“Even before CMU, I was working on autonomous ground vehicles and driverless cars, and had a lot of experience in autonomous navigation,” Galluzzo said. “But then at CMU we got to work on a project for the department of defense to do autonomous manipulation. We surprised ourselves how well we could do certain tasks, particularly when it came to picking up objects and moving them around.”
Galluzzo went looking for the low-hanging fruit in industry where he could find applications for autonomous manipulation. The startup spent its first year conducting extensive market research.
“We talked to a lot of executive level folks in the industry and got a lot of validation. Everyone we talked to said, ‘Yes, we want this product yesterday!’" Galluzzo said. "This has been a dream of ours for 5 years, but the industry has been dreaming about this for decades, putting robot arms on mobile AGVs and having them do useful, everyday work inside a distribution center. Now it’s here.”
Further validation came with VC investment by Comet Labs, an intelligent machine and robotics accelerator based in Silicon Valley. Still in the seed-funding stage, IAM Robotics is accepting pilot program customers and launching initial beta deployments. They have a preferred partner program to provide additional incentives for early adopters.
“We have a leasing partner that’s working with us to finance robots for customers on a case-by-case basis," Galluzzo said. "It makes for a very fast return on investment, especially when you’re building robots that can work as fast as a person. Our first customer is Rochester Drug Cooperative (RDC), so we’ll be actually deploying the system into a pharmaceuticals facility this summer."
Other potential applications for the robot include warehouse distribution and order fulfillment in the health and beauty, grocery, and e-commerce markets.
IAM Robotics’ technology uses a depth-sensing camera for navigation. For maneuverability it uses two center-drive wheels with casters around the outside, which are all covered for safety. It has a swappable rechargeable battery that lasts about 10 hours, so the system can operate 24/7 by simply exchanging spent batteries for fully charged ones. The onboard HMI display shows battery level warnings, along with other critical status elements.
Collaborative mobile picking robot
The robotic manipulator is a FANUC LR Mate 200iD six-axis arm guided by the same vision technology as the mobile base. Swift uses the same depth-sensing cameras to inspect the surrounding environment for collision avoidance. Both the mobile base and the robot arm will slow down and then stop if they sense the presence of a person or other obstacle impeding their safe movement.
The system relies on a patent-pending 3-D product scanner called Flash, also created by IAM Robotics, that records the barcode, product dimensions, weight, and 3-D features of the items intended for picking. The picking robot must first “learn” what each item looks like and how to grasp it. Then it uses onboard software called RapidVision that enables Swift to recognize the items it’s trying to pick.
“The intelligence of the system and the ability for it to be autonomous is all driven by software,” Galluzzo said. “The system actually connects to remote tablets and mobile devices, so that it can send data to those devices and let people know that it needs assistance.
“We’re not taking the people completely out of the loop,” he adds. “We’re just trying to keep the time between human interventions minimal.”
The shelves also have to be stocked in a fairly organized fashion, so that the robots can securely grasp the items.
“We divide and conquer the warehouse,” Galluzzo said. “We look for warehouses that have a high percentage of inventory that is compatible with the robot. We have people doing what they’re good at, which is handling the deformable items like apparel, and we have the robots doing what they are good at, which is picking the rigid boxes and bottles.
“Ergonomic advantages are key on our customers’ minds in this market. Because of the exponential expansion of e-commerce, there just aren’t enough people to do the work. Because the robot is doing all the lifting, it makes the job easier. It’s bending down, it’s reaching up high, and it’s doing this all day long. Now a person is upgraded to not just a picker, but a supervisor of robots.”
By automating the picking and transport processes, these mobile robots have the ability to collect video and pictures, and track movement of product. Again, ready for Industrie 4.0.
“By having machines that can basically work as data collectors, connected devices in a manufacturing and distribution facility, you now have an unprecedented ability to collect and manage data which you didn’t have when you just had people walking around,” Galluzzo said. “Now we can almost turn an entire facility into a video game-style command and control center. Rather than taking 5 to10 minutes to walk through my facility to see if something is on the shelf, I can just pull up a picture from the live robot view of the world.”
Large-payload mobile manipulators
A new breed of free-roaming mobile manipulators will be able to pick and transport items in a range of sizes and payloads. They are becoming the new face of warehouse logistics and order fulfillment.
At Automatica in June, Otto Motors and Japanese robot maker Yaskawa Motoman teamed up to unveil their R&D effort for combining the larger Otto 1500 mobile vehicle with an onboard six-axis robot.
Bastian Solutions is working with the Southwest Research Institute (SwRI) to develop its own brand of mobile manipulators. Joe Zoghzoghy said the technology is finally mature enough to build these kinds of systems. At the same time, demand for mobile manipulators is rising as competition for faster cycle times and higher throughput intensifies between distribution and fulfillment centers.
“What differentiates these systems from other mobile robots and shuttle systems is their ability to fill multiple orders at a time,” Zoghzoghy said. “We built the (autonomous) AGV system in house. The arm we get from robot suppliers. Everything else on board is custom designed ourselves.”
He said the systems are running on ROS open-source software and Bastian is working closely with ROS developers at SwRI. The open source software platform means greater flexibility for design changes, which can be a tremendous advantage.
“If we need to go with a different arm for whatever the reason is, say we need to have a bigger robot or a client has a preference for a certain robot manufacturer, we can swap out arms without going through the software development for that specific supplier,” Zoghzoghy said. “ROS gives us modularity in a sense, so that we can use any type of mechanical, electrical, or electronic system and integrate it for our solution. The key term in the past has always been goods-to-person. We’re thinking the next logical step will be to make the robot (manipulator) go to the goods, much the way a person does.”
With the dawn of “robot-to-goods,” the face of intralogistics is transforming before our eyes. The swarm effect will take hold with free-roaming mobile robots as more manufacturers upgrade their automation and improve their supply chain.
Tanya M. Anandan is contributing editor for the Robotic Industries Association (RIA) and Robotics Online. RIA is a not-for-profit trade association dedicated to improving the regional, national and global competitiveness of the North American manufacturing and service sectors through robotics and related automation. This article originally appeared on the RIA website. The RIA is a part of the Association for Advancing Automation (A3). A3 is a CFE Media content partner. Edited by Chris Vavra, production editor, Control Engineering, CFE Media, email@example.com.
See additional stories about ROS and other robotics developments linked below.
For more information about upcoming A3 events including A3 Business Forum and Automate, visit: www.robotics.org.
See a related article about robotic telepresence linked below.