Every returned item has to be picked up, identified and sorted before it can be sold again, and that work has long resisted automation.
“If a system only works on articles it has seen before, it does not work in a returns operation at all,” Ralf Gulde, co-founder and CEO of German robotics software company Sereact, said in an Oct. 7 announcement of a new artificial intelligence-powered returns processing system.
Robots powered by Sereact’s AI are now sorting returned clothing for Berlin-based online fashion retailer Zalando, with 62 million active customers. CEVA Logistics, the Marseille-based logistics arm of French shipping group CMA CGM, runs the robots at fulfillment sites in Greven, Germany, and Świebodzin, Poland.
Sereact’s AI lets the robots handle objects they have never seen. The three companies called it the first phase of a plan to scale returns automation across Europe.
Sereact Robots Sort Clothes Without Item-by-Item Training
The dual-arm robots grasp, identify and sort returned fashion items without article-specific training or predetermined article sets, the companies said. They run on Cortex, Sereact’s AI platform. CEVA delivers the system to Zalando as Robotics-as-a-Service.
With the robots taking over repetitive handling, warehouse teams can shift to station supervision, exception management and quality control, the companies said. Chris Walton, senior vice president of contract logistics at CEVA, called returns handling “one of the toughest challenges” in logistics. He said the service classifies items faster, cuts manual touchpoints and keeps quality consistent at scale.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
Zalando joined Sereact’s $116 million Series B round as a strategic investor in July. Sereact’s other customers include IKEA, DHL, BMW and Mercedes-Benz, according to the release.
UPS, Amazon and Walmart Automate Around the Return
In the U.S., UPS-owned Happy Returns has automated parts of returns sorting. New technology, including Geekplus robots, lifted its item sorting efficiency by 50%, UPS said. Happy Returns deployed 150 Geekplus sorting robots at a returns hub in eastern Pennsylvania, according to a Geekplus case study. In September, Happy Returns launched box-free, label-free returns for TikTok Shop sellers. Shoppers can drop off items at nearly 10,000 Return Bar locations, and the returns move through the UPS network.
At Amazon, returns go to centers dedicated to product types such as clothing, electronics and furniture. Workers there check each item before it is resold, sent back to sellers, liquidated or donated, the company said. Items that no longer qualify as new are graded into four conditions and sold through Amazon Resale. Amazon uses AI earlier in the chain. Its Project P.I. system combines generative AI and computer vision to flag damaged or wrong items in fulfillment centers before they ship, the company said.
Walmart has focused its automation further upstream, on the warehouses that supply its stores and online orders. About 60% of its U.S. stores receive freight from automated distribution centers, and about half of its U.S. eCommerce fulfillment center volume is automated, PYMNTS reported in March.
58% of Merchants See AI Agents Handling Returns
Merchants expect AI to take on more of the returns process. Among 1,185 merchants in the U.S., Brazil and the UAE, 58% agree at least somewhat that AI agents will automate returns, refunds and issue resolution, according to the PYMNTS Intelligence report “The Agentic Commerce Deep Dive: Payment Infrastructure and the Path to Trust for Merchants.”
Merchants also want to keep control. In the same report, 38% named return and refund rules as a non-negotiable area where they need to retain control over AI agents. Shoppers are more open. Among 5,241 consumers surveyed, 46% said they would let an AI agent initiate returns for them.
Add as Preferred Source
