|Automated warehouse picking involves collaborative effort and is not only reserved for major enterprises endowed with unlimited budgets or resources / Photo by: Pisit Khambubpha via 123RF|
It can be daunting to develop a picking technique that caters to the needs of warehouses, especially if they are dealing with challenges like having a diverse inventory, a revolving door pf pickers, a number of SKUs, or even trying to achieve lightning-fast picking goals, said Fergal Glynn of 6River, a warehouse management and automation firm. However, the savviest and most future-oriented of all warehouses are investing in automation to streamline the picking process.
Automated warehouse picking involves collaborative effort and is not only reserved for major enterprises endowed with unlimited budgets or resources. The best automated solutions can be tailor-made to add value to a warehouse’s existing picking processes without the need for downtime or waiting to achieve ROI. But there’s even greater news for warehouses that want to further optimize their picking processes: an AI-powered warehouse picking robot.
Warehouse Automation Statistics
Research firm LogisticsIQ estimated that the global warehouse automation market will grow to more than twice from $13 billion in 2018 to $27 billion by 2025 at CAGR of 11.7% between the forecast period 2019 to 2025, as cited by robotics news website Robotics Business Review. LogisticsIQ also found that the AGV (automated guided vehicles)/AMR (autonomous mobile robots) market is projected to reach the $4 billion mark by 2025, with AMRs being the main contributor in retail warehouses.
This is due to AMRs having a high demand in the e-commerce industry as well as its flexibility to deploy the unit without altering the existing warehouse’s structure. Nevertheless, the AGVs/AMRs market is forecasted to have more than 15% market share in overall warehouse automation, which will be led by key players like Dematic, SSI-Schaefer, Geek+, GreyOrange, and HikVision.
Researchers from Westernacher, a provider of global business and SAP consulting, predicted that the number of robots deployed in warehouses will grow 15 times (600,000 units) by the end of 2021. Over 10% of US warehouses were using automated warehousing equipment that enables a goods-to-person picking approach in 2016. Those figures are said to skyrocket in the coming decades.
Roberto Michel of logistics news platform Logistics Management wrote that manual picking was used by 72% of warehouses in 2019, a 4% decrease from 2018, cited Ed Romaine of Convey Co, a material handling systems integrator. This reflects the changing dynamics in the order fulfillment industry, including a tighter labor force and space restrictions. These are the factors that help drive technological advances in today’s modern warehouses.
What Is Automated Warehouse Picking?
An automated warehouse is defined as implementing robotic or semi-robotic technologies to augment the work of human pickers. Warehouses have a wide range of options to choose from when it comes to automated warehouse picking. However, choosing the most efficient automation tools to help warehouse staff can be easily employed into existing warehouse operations quickly and seamlessly.
Automated warehouse picking can also slash walking time and shorten picking routes. It can also integrate with existing WMS and support accurate picking and packing of goods. Automated warehouse picking can also be employed in various industries such as e-commerce, manufacturing, transportation, food and beverage, medical equipment, retail, and more.
Automated warehouse picking is flexible, offering warehouses a more customizable approach than traditional warehouse automation systems. Unlike modern warehouse automation tools, traditional systems are bulky and consume warehouse spaces.
|An automated warehouse is defined as implementing robotic or semi-robotic technologies to augment the work of human pickers. Warehouses have a wide range of options to choose from when it comes to automated warehouse picking / Photo by: macor via 123RF|
Great News! AI-Powered Robot Warehouse Pickers Are Here
Covariant came out and announced its partnership with KNAPP, reported Karen Hao of MIT Technology Review, a tech magazine. The company’s algorithms have been deployed on KNAPP’s robots in two of KNAPP’s customers’ warehouses. This is a game-changer in the field of AI-driven robotics, in which systems were in highly-constrained academic environments. But Covariant said it is ready to deploy its system to warehouse floors.
Robots have been in warehouses for a long time, but their success has been limited to moving boxes from the front to the back of warehouses. Peter Chen, co-founder and CEO of Covariant explained, “If you look at a modern warehouse, people actually rarely move.” However, moving goods with hands requires more than having the right hardware. The system needs to adapt to a variety of product shapes, sizes, and their ever-changing orientations.
A traditional robotic arm can execute the same movements repeatedly, but it will fail once there are any deviations. Therefore, the robotic arms needs AI to “see” and adjust to keep with any changes to its surroundings. In order for warehouses to get their money’s worth, a robotic arm needs to be at least 99% and perhaps 99.5% accurate in order for it to operate with little human intervention or risk of slowing down a production line.
And Covariant has done research on that. Members of Covariant’s team visit convenience stores to purchase items from medicines and packaged clothes to eraser caps inside clear boxes. They also looked for objects that might trip up the robotic picker like bags of chips, transparent plastic surfaces, and more.
A camera that acts as the robots’ eyes hangs above them. The visual and sensor data from the robot’s body is fed into the algorithm that helps control its movements. The machines learn through a combination of imitation and reinforcement techniques. First, a person manually guides the robot to pick up various objects. The robot then logs and analyzes the motion sequences to generalize the behavior— which involves millions of rounds of trial and error. When the robot reaches an object, it tries to do it in slightly different ways, logging the attempts that result in faster and more precise picks. This aids the robot in improving its performance.
In just 60 minutes, three robots masterfully picked up all sorts of store-bought goods. In seconds, the algorithm analyzes the goods’ positions, calculates the angle and correct sequence of motions, and moves its arm to grab the object with a suction cup. The robots move with certainty and precision. They can also change their speed based on the delicateness of the product.
KNAPP’S Puchwein noted that its robots have picked up between 10 and 15% to 95% of the product range of Obeta, a German electrical supplier. The remaining 5% is alotted for fragile items such as glasses, which are still done by humans. Puchwein stated, “In the future, a typical setup should be maybe you have 10 robots and one manual picking station.”
Warehouse automation systems like robot pickers are more flexible than traditional automation systems. However, they’re not good at handling slight deviations in repetitive movements. Fortunately, Convariant developed an AI-powered robot picker that could pick up a variety of goods with precision. AI-enabled robot pickers will hopefully encourage more warehouses to invest in these units to augment their picking operations.