Faster, please, with no mistakes. That's what most logistics managers are being asked to do with respect to the movement of products through their distribution centers. Companies today want to push products through their facilities more rapidly than ever without incurring higher costs or sacrificing accuracy.
That's a tough assignment. But one way a logistics manager can increase warehouse throughput is by deploying camera-based data capture in conjunction with analytics. Cameras snap a series of pictures of packages moving down a conveyor. Those images can be stored in a database and then retrieved by software for analysis of the operation. Companies marketing these type of systems include Datalogic and Sick Inc.
Searching a database of images for operational insights is the latest example of "big data" analysis applied to the supply chain. All those pictures would be just a pile of big data if it weren't for the development of special software that can filter and search through the images. "You're sifting through millions and millions of images to look for something significant," explains Mark Kremer, director of sales for retail logistics at Sick. "You can use filtering rules to look for a specific set of images."
Analysis of the images can be used to pinpoint problems in a distribution operation so that corrective action can be taken. For example, the system might identify a too-large gap between packages moving on a conveyor. By shrinking the gap from, say, 16 inches to 12, a distribution center could speed up the flow, increasing throughput.
Analysis of the images could also be useful in determining whether suppliers are in "bar code compliance"—that is, whether they're meeting their customers' specifications for how and where bar codes should be placed on a package or case. Instead of having a worker at a retailer's DC conduct a manual audit for bar code compliance, the software could perform the audit by searching the database.
In addition, an image analysis could be used to determine the root cause of "side by sides" in sortation systems. Side by sides occur when a smaller package gets squeezed up against another, larger package, resulting in the two products' being mistaken for a single unit. "The camera system analyzes the image to determine when there is a condition of multiple cartons coming down the conveyor," says John Park, a marketing product manager with Datalogic Automation. By reviewing images of "side-by-side" packages, managers can figure out what's going on and make adjustments to correct the problem.
As more companies get involved in supporting online sales, this type of big data analysis could also be used by customer service for product verification, Kremer says. For example, as a warehouse worker packs a customer order—say, a pair of boots—a picture could be taken and stored in the database. When the customer calls up and complains that the right boots didn't arrive, a customer service rep could call up that image to determine whether the correct merchandise was shipped.
A supply chain manager could also use these types of analytics to compare DC operations in the network, utilizing the software to conduct an image analysis for each warehouse and generate specific reports on each facility. The manager could use these reports to determine why one DC has a higher throughput rate than another. "You could use this tool to compare DCs across the globe," says Kremer.