Machine vision technologies boost warehouse efficiency, transparency
The machine vision industry is providing warehouses with technology to leverage the data generated by cameras, sensors and imagers across the entire enterprise in order to maximize efficiency and productivity.
Amazon has gotten warehouse efficiency down to a science, and vision and imaging technologies share in the credit. Meanwhile, human pickers are equipped with handheld image-based barcode readers, and merchandise gets scanned at a series of points throughout the fulfillment process.
It’s not just retail giants, however. Warehouses of all types and sizes are realizing they don’t have to be an e-commerce gorilla to benefit from vision and imaging offerings ranging from 3-D empty tray detection to smartphone-based scanners. Not only is the vision industry providing the hardware, but they’re helping their warehouse customers leverage the data generated by cameras, sensors and imagers across the entire enterprise in order to maximize efficiency and productivity.
Moving beyond the barcode
Barcode readers continue to be the bread and butter of imaging applications in the warehouse. These industrial imagers may require less complexity than their traditional machine vision system counterparts, but they nevertheless have a demanding job.
"Compared to factory automation and its many applications, the warehouse is more focused and contained because you are just trying to read a barcode and guide a box where to go," said Bryan Boatner, director of product marketing -mobile and handheld products for Cognex. "But in another sense, because the boxes are moving so fast and throughput is at such a premium, it can be a lot more challenging."
When image-based industrial barcode readers first debuted, their primary value proposition was providing better read rates and capturing more data than their laser scanner counterparts ever could. In reading direct part mark (DPM), 1-D, and 2-D barcodes, image-based devices allowed warehouses to save images of codes that couldn’t be read in order to perform troubleshooting and root cause analysis to help improve the process.
Once customers saw that benefit, they started to explore ways to utilize other information available from the barcode image. Until recently, much of that data was tossed aside, said Bradley Weber, manufacturing industry product specialist and application engineering manager at Datalogic.
"It used to be taking that image, running it through algorithms such as reading the barcode, and processing it right then and there, and then moving onto the next package," Weber said. "It’s changing now where a lot of that information is being stored and analyzed for later so you can identify trends over time."
To help keep its customers’ data from languishing, Cognex developed the Cognex Explorer real time monitoring (RTM) system. When an unread barcode is detected, it automatically transfers the image to RTM, which is designed to evaluate each image and categorize them into a group based on their error—for example, missing labels and poorly printed labels. Categorized images are stored in a database accessible via a web browser. Just as warehouses are relying on the additional data generated by barcode readers, they’re embracing a multipart vision system to track a product from the front end to the back end.
"Datalogic is helping our customers to build a fingerprint of the package traveling through the facility using different vision technologies, including barcode readers that integrate optical character recognition (OCR) functionality, sensors that detect the presence or absence of a package, dimensioners that scan the package to provide its volume, and machine vision cameras capturing what is on that package," Weber said.
For its part, Cognex is investigating ways to employ its vision technology on its handheld readers and mobile terminals to automate OCR. "We plan to demonstrate to customers how they can use vision on handheld readers to read ZIP codes off the label in addition to scanning the bin location barcode," Boatner said. "You can even conceive the ability to do the entire form reading where you convert printed fields to an automated data collection service."
Image-based warehouse management machine vision systems are enabling companies to be more transparent—and therefore better partners—to their customers. Datalogic’s Weber says that all information gathered together from dimensioner, weight, images, or other sensors can be combined to have a unique ID for a package. "The customers will be able to access it all," Weber said.
Shop floor to top floor
Delving this deep into the data, of course, depends on the ability of the manufacturing plant or distribution center to connect disparate business activities, enabling dataflow for centralized decision-making. But many of these systems, including warehouse management, continue to operate in silos. To tie warehouse activities into front office operations—colloquially known as "shop floor to top floor"—vision products and systems ideally will integrate with an enterprise resource planning (ERP) system, which centrally manages an organization’s business activities and the data generated by them.
But an ERP comes with its own challenges. "The ERP does just enough to get by, and it’s not very nimble," said Dan Hare, vice president of Matrix SSI, which provides technology solutions to centralize inventory control. "It doesn’t do a good job at device management, which is part of the inventory control because you’re printing labels and scanning. We understand the workflow that happens in the warehouse better."
Of particular focus for manufacturers is automatically tracking work-in-process inventory, or the raw materials that are being transformed into finished goods. "A lot of this is being done with serialization and lot-type tracking, and older systems just don’t do a good job of that," Hare said. "We’ve seen a big uptick in our solution used as an overlay for [ERP systems such as] SAP and Oracle, and that’s even in brand-new installations."
And without being responsive to the needs of the warehouse worker, workflow is likely to suffer. "Many ERPs are designed for big computer screens with a mouse at a desk," Hare said. "Matrix SSI is designed to run on handhelds on the shop floor."
Matrix SSI, which is hardware-agnostic, has spent the past 15 years building an integration tool to connect all silos and disparate systems in a warehouse. It’s an engineering effort that vision companies are starting to make for their customers as well. Cognex’s RTM, for example, is designed to harness all the data it collects on images, which can be distilled and presented to office managers so they can assess the information and make changes.
The vision-enabled mobile terminal employs technology used across the entire enterprise and leverages a variety of Android and iOS smartphones as the user interface of the device. The phones are set within a ruggedized housing equipped with barcode reading algorithms. Boatner said, "It’s a much easier solution to deploy because you have one device that is managed by IT that is deployed in your warehouse, front office, field team, and so on."
Today, the machine vision industry is chasing every opportunity to catch the "killer app," which is usually associated with the massive installed base of smartphones around the world. The warehouse—with its controlled, demanding environment—may be the gateway that brings machine vision technology to everyone’s pocket, whether they are a distribution center manager or a customer.
Winn Hardin is contributing editor, AIA. This article originally appeared on the AIA website. The AIA 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, firstname.lastname@example.org.
See additional stories from A3 linked below.
Original content can be found at www.visiononline.org.