German clothing store chain Adler Modemärkte is expanding its use of RFID technology with a tag-reading robotic system that now tracks inventory at 20 of its stores, with plans to deploy the system across 45 locations—about one-third of the retailer’s sites—by September 2019. The system consists of a robot known as TORY, which has a built-in UHF RFID reader and antenna array that collects RFID data during stock-taking trips throughout the stores. The technology was provided by RFID technology company MetraLabs (see MetraLabs’ TORY RFID Inventory Robot Celebrates First Jubilee).
Adler is one of the largest retail chains in Germany, with stores throughout that country, as well as in Austria, Luxembourg and Switzerland. It mostly sells its own brands for women, men and children, though around 25 percent of its goods come from external brands. For years, it has applied UHF RFID tags to its branded products and incoming external brands in order to better manage inventory counts as items move through its distribution center and stores, thereby reducing the risk of goods going out of stock at any specific site. The tags were periodically read at the stores via handheld readers. Workers walked throughout each of the 177 stores on a weekly basis to capture the EPC UHF RFID tags, enabling the company to regularly update its inventory count.
With the RFID deployment, Adler’s goal has been to increase its inventory accuracy and avoid the misplacement of goods within its stores. While the company has used its staff to track inventory and return goods to the proper place on store shelves and displays, this takes them away from their primary task of sales and assisting customers, says Patrick Schiller, Adler Modemärkte’s CIO.
The company began working with MetraLabs in 2015, at which time it started testing TORY’s ability to take on sales associates’ task of scanning tags to identify the presence and location of store merchandise. The robot piloting took place at a handful of stores (see German Clothing Retailer Adler Gives RFID Robots a Spin).
“During the pilot phase, the physical challenges of RFID technology were the primary consideration,” Schiller says. The first version of TORY, with its 11 antennas and a reading range of more than 7 meters (23 feet)—up to 30 feet—is common, he says. That version could read tags through walls. As such, the software algorithms for stock counts had to be regularly adjusted. For this reason, the retailer has painted the storage area’s walls with RFID-shielding paint.
The TORY robot comes with an off-the-shelf UHF RFID reader, an array of patch antennas developed in house, and laser and camera sensors to identify its path and current location. Its diameter is 50 centimeters (19.7 inches), while the antenna tower in the middle can extend to more than 7 feet. The robot’s typical reading speed is approximately 0.5 to 1 meter (1.6 to 3.3 feet) and about 250 tags per second, which is normal walking speed. However, that can depend on how many tags it sees at any given time, says Johannes Trabert, MetraLabs’ co-founder and executive partner. Its built-in memory can store more than one million tag reads. The robot forwards the data it collects to a server via a Wi-Fi or wired connection.
According to Trabert, the antenna array was designed “to yield high accuracy and read rates in typical retail scenarios—both very tall and low shelves, stuffed boxes of merchandise, and multi-path-propagation problems due to metal shelving.” While the initial version of the robot had an array of 11 antennas, the latest iteration has 16, thus expanding the read area, as well as making reading faster and more sensitive.
To launch the inventory-tracking functionality, a user at each store sets up the robot’s route or coverage area. This is accomplished using a temporarily attached touchscreen on the robot, or a remote PC connection via Wi-Fi, to communicate with its onboard software. The software displays instructions on the touchscreen to guide users through the process of setting up a new “scan area.”
The retailer’s own database software stores data indicating what inventory should be located at each store shelf or on each rack. If the robot fails to capture the tag ID numbers of the items expected at a given location, the device can return to that section to attempt another reading. It can also circle back for a second read in areas where tagged items are densely packed together, meaning the likelihood of a missed tag read is high, then forward data about any missing items via Wi-Fi. The robot accomplishes its inventory-counting tasks approximately 10 times faster than a manual count and more accurately than with a handheld reader, Trabert reports, based on user trials carried out with retailers.
The transition from pilot to rollout at scores of stores, Schiller says, has required modifications to manage the large amount of inventory data the robot is collecting. “The challenges during rollout are in the data volumes that are generated every night and have to be processed and booked in the RFID system and the ERP [enterprise resource planning system].” At the TORY-operated stores, he says, “We make a full inventory every night, [which] we did once a year in the past.”
One of Adler’s main goals, beyond better inventory visibility, is to reduce the incidence of out-of-stock events, Schiller says. In addition, the company hopes to gain analytics from the solution that will help it with future planning. “We will also be able to track the movement of articles through the shops and thus optimize space,” he states.
The retailer plans to equip all 175 shops with a TORY robot by 2021. “Our goal is to further increase inventory security and avoid misplacements,” Schiller explains. “In addition, our employees can use the additional time freed up for sales. This was also the basis for the business case.” By using the TORY system, he notes, the company can measurably reduce inventory disparities. “Due to the large areas of our shops, we can generate targeted replenishment suggestions with the RFID search.” According to Trabert, Adler is rolling out the system at a rate of two new robots every week.