SDI, a California-based logistics OEM and integrator, is responsible for designing and installing machine vision systems for its customers, including leading fashion retailers like Nordstrom, Lululemon, and Levi Strauss. As the demand for quick online deliveries continues to grow, the stakes have become even higher for order accuracy. For example, when a single item is ordered by a consumer, but the wrong product is delivered, it can result in a very costly mistake for retailers.
For SDI, reducing and preventing the amount of false or incorrect detections of products was an area they hoped to help their customers improve upon. To do that, they needed the right machine vision solution to increase sortation accuracy at customer sites.
“For store fulfillment, if we mis-sort an item or item detection doesn’t detect something properly and it gets sorted to the wrong place, it’s an aggravation, but not a tremendous issue because the item just goes to a different store,” said Dave Bodenheimer, VP, Research and Product Development at SDI. “In today’s world, we’re moving more into order fulfillment and in these cases, mis-sorts are incredibly difficult and painful for the client. So, the role that item detection plays now is becoming much more important.”
Initially, SDI used a photo eye solution for item detection on unit sorters for its customers but needed something more robust. As technology continued to advance, SDI was introduced to Cognex’s latest logistics innovation, the In-Sight 2800 Detector. Using AI-based edge learning technology, the Detector can uplevel key applications within sortation, including item detection, item classification, and process issue detection.
“We anticipate that these newer solutions from Cognex, like the In-Sight 2800 Detector, are going to give us the tools we need to bring that accuracy up to a level that the customer will be happy with.”
For example, the Detector helps SDI’s customers detect presence or absence of items on the most challenging backgrounds and in low-contrast scenarios. The Detector also allows retailers to distinguish between any number of packaging types including boxes, polybags, jiffy bags, paper bags, and more. And knowing that many of SDI’s customers use different variations of sorters, the Detector makes it easy to identify process issues like conveyor jams or hygiene issues like dirty totes.
SDI upgraded one customer to the Detector to improve presence/absence detection on its sorting machines. The ease of setup of the Detector was critical for the retailer; SDI was able to set up 10 cameras in just 4 hours. “Edge learning helped us in terms of the speed of setup,” said Bodenheimer. “In fact, we were able to get a baseline on one of our sorters by simply training one of the cameras and moving that training across to the other cameras.”
Now, SDI is working with numerous customers to switch their sortation solutions to the Detector. “With the In-Sight 2800 Detector, we were able to immediately see products that were not able to be detected by other solutions,” he said. “Our customers have been very pleased with the results. The In-Sight 2800 Detector is a game-changer for presence/absence, detection, and simple classification.”
For more information, watch our customer success story video with SDI.