The Robot Uprising Reaches the Store

robot uprisingThe robot uprising is coming to a store near you. And no, I’m not talking about Skynet building T-1000 Terminators to take over the world (or store in this case). I am talking about how the retail store is drastically changing to become more automated, using artificial intelligence, and relying on non-human assistance for check-outs, price checks, and inventory management. Robots and drones are in stores now and retailers are looking to build upon their capabilities for the foreseeable future.

The Robot Uprising

Robots in the warehouse are certainly nothing new. My colleagues Clint Reiser and Steve Banker have written extensively on how robots and automation are changing and shaping the warehouse of the future. Autonomous mobile robots are coming of age, as they get smarter and enable workers to be more productive due to constant collaboration. Two examples of these bots are Locus Robotics and Vecna Robotics. The Locus solution coordinates autonomous mobile robots to work collaboratively with warehouse workers. The bots dynamically navigate the warehouse floor meeting up with manual pickers until the tote is full and then it and transports the full tote to the pack station, and subsequently returns to pick zones for the next round of work. In contrast, Vecna Robotics‘ Tote Retrieval System uses vacuum technology and an extending tray to extract boxes and totes from warehouse shelving onto a mobile robotic cart.

Robots have also garnered a lot of attention around home delivery. Companies like Starship Technologies have developed autonomous mobile robots to complete the last mile of home delivery. There have been a number of pilots run by retailers of all shapes and sizes, but Starship has actually moved beyond the pilot phase of home deliveries and has launched an on-demand package delivery system. The system requires customers to download a mobile app, which gives an alternate delivery address to go in the place of their home address. Once the package arrives there, the app will notify the customer and enable them to request a Starship bot to deliver it to them, wherever they are. Through the app, customers can also track where the package is at all times. Starship is also expanding its delivery program to universities.

But robots are no longer just for warehouses or home deliveries, as more retail stores are deploying bots for a number of functions. Walmart and Stop & Shop are just two examples of stores that are using robots within the four walls of the brick and mortar store. Walmart has deployed Bossa Nova robots at more than 350 of its stores nationwide. These autonomous robots roam the store aisles checking for pricing issues, out-of-stocks, and shelf irregularities. As Steve Banker wrote, these robots are mostly used for real-time, on-shelf product data. Many robots have a difficult time in the store counting on-hand inventory or even recognizing inventory that may be slightly askew on the shelf. The Bossa Nova robot is not designed for this practice. Instead, it is used to identify whether a slot is empty or full, and identify which SKU is in a slot and the price associated with that slot.

Stop & Shop has a robot named Marty that also wanders the aisles of its stores. These robots are not designed for inventory management or stock-out identification. Instead, Marty is designed to look for hazards in the store, whether that is a spill, trash, or fallen signs. When Marty detects a hazard, it takes a picture, sends it to a central hub for confirmation, and then alerts customers and employees (in English and Spanish) of the hazard. Employees can then address the hazard.

One of the biggest obstacles to rolling out robots in the store is public perception around safety. These robots have embedded cameras and sensors, as well as intelligence engines. These technologies should help them to navigate the stores safely, avoiding customers, employees, displays, and other obstacles. However, all it takes is one misstep and the public backlash can be unforgiving. Privacy concerns also abound, as customers are clearly curious to see if these robots will use cameras to take their pictures to match customers to buying patterns. Either way, more and more retailers are investing in robots in the stores for inventory management and customer service practices.

The Drone Uprising

When it comes to drones, the last few years have created a lot of hype around last mile deliveries. However, previous FAA regulations essentially made drone deliveries an impossibility. That didn’t stop companies from investing significant R&D dollars into drones for home deliveries, with different models created for different delivery scenarios. However, in late 2017, President Trump announced plans to ease drone regulations and increase test flights to make the use of drones a possibility for home deliveries. The FAA has since approved a number of pilot programs for drone initiatives, some of which will likely include home delivery. There are still a number of challenges facing drone operations in the United States. As an aside, the single biggest application of commercial drone use is still real-estate aerial photography.

From an application standpoint, home delivery does not come to the top of my list of practical uses. Instead, I look at three main categories for drone usage: medical supply delivery, ship re-stocks at ports, and inventory management. Nearly three years ago, Zipline began making using delivery drones for moving medical supplies to health clinics in Rwanda. These supplies include blood, medical equipment, and medicine. Maersk has run numerous tests to stock ships at sea without having to use smaller ships to bring the supplies. The goal is to reduce the cost of restocking ships by using the drones. And finally, PINC Solutions provides drones for use in yards and warehouses that make use of RFID for real-time location of yard assets and inventory management processes within the warehouse.

Inventory management has always been the bane of brick and mortar retail’s existence. As omni-channel practices have become pervasive, it is has become more apparent that there are problems with inventory accuracy, and the store is seen as the weak link when it comes to omni-channel. Well, I recently came across a company that is looking to use drones for inventory management in the physical store. Having drones flying around the store seems like a recipe for disaster, but Pensa Systems is using their drones to tackle inventory stock-outs.

Pensa is based in Austin, TX, and has developed a drone for in-store inventory management. The drones are small and quiet, and have the rotors enclosed in a protective housing to make them safer. The key to using the drones is to deploy them after hours or selectively during store hours when areas are free. Using free space is clearly the preferred method, as after-hours usage will not identify how long a shelf has been depleted of merchandise.

The drones use computer vision and artificial intelligence for a real-time analysis of shelf conditions. The camera on the drone is able to capture real-time shelf conditions and feed the data into an AI engine to recognize how products are managed, but also to predict future patterns for stock-outs.

Other companies have looked into the use of drones within the store. Walmart, for example, filed a patent to develop a customer service drone within the store. The drone would communicate with a customer’s mobile device to help locate products or run price checks. However, the system has not been rolled out yet.

Final Thought

The robot uprising is clearly underway. But according to manufacturers and end users, the end goal is not to eliminate human workers in the store. Instead, the robots are there to assist humans and make their jobs easier. Additionally, robots and drones are designed to eliminate tedious, manual processes, freeing workers up to accomplish other tasks. While the warehouse has been at the forefront of automation and robot adoption, the store is gaining ground as it tries to become more efficient.

Leave a Reply

Your email address will not be published. Required fields are marked *