Pull up a truck to the dock door of a DC in 2021, and the lift truck that arrives to unload your pallets might have a driver behind the wheel ... or it might have a bundle of sensors. Inside the building, the hand reaching into a tote to retrieve an item for an order might be connected to a human laborer … or it might be attached to a mechanized arm. Over at the racks of stored goods, the bar-code scanner taking inventory might be wielded by a warehouse employee … or it might be mounted on a hovering drone.
Logistics robots are here to stay, and they're whirring around every corner of the DC, helping companies handle the surge of e-commerce orders triggered by the pandemic. But demand for warehouse robots began long before the coronavirus reached U.S. shores in 2020. Companies have been eyeing the technology for years, drawn by its potential on a number of fronts. In particular, they've been looking at robots as a way to compensate for labor shortages and help them train temp workers during peak season, match Amazon's shipping speed, and pack more inventory into their facilities.
All robots are not the same, though. Like the animals on a farm, bots come in all combinations of shape, size, speed, strength, and smarts. This primer on robotic technology can help "farmers" determine which of the many options best meets their needs.
An automated storage and retrieval system (AS/RS) can automate many warehouse processes by storing, delivering, tracking, and replenishing inventory through a computer-controlled process that automatically deposits and retrieves loads from set storage locations in a set of steel racks.
Installing an AS/RS is a long-term investment, requiring a substantial upfront cost, a dedicated area of the warehouse, and specialized racks and totes for many models. Furthermore, because each unit is designed for goods of a specific size and shape, these systems lack the flexibility to handle variation.
But once the unit is up and running, an AS/RS can deliver high throughput speeds, reducing the size of the workforce needed to move goods around the facility, eliminating fulfillment errors, and easily coping with inventory challenges like high SKU (stock-keeping unit) counts, high-value goods, or heavy items.
These AS/RS solutions come in many varieties, including models with cranes reaching between aisles to fetch racked goods, units with shuttles that glide above the storage racks or whir between them in three dimensions, and vertical storage carousels and vertical lift modules that store goods in a self-contained unit.
Automated guided vehicles (AGVs) have roamed warehouse aisles for decades, carrying inventory along set routes demarcated by wires or magnets embedded in the concrete floor, bar-code stickers affixed to storage racks, or wireless signal beacons mounted on walls.
That system allows vehicles like self-driving tugs, forklifts, and pallet jacks to steer themselves between predetermined indoor locations and shuttle goods from point to point. Recent upgrades have added sensors like computer-vision cameras or LiDAR (light detection and ranging) technology to enhance their ability to detect obstacles and avoid collisions. But AGVs currently don't have the capability to change their routes, adapt to new workflows, or communicate with other vehicles to optimize the movement of materials.
However, vendors say the lines between AGVs and their cousins, autonomous mobile robots (AMRs), are beginning to blur. For example, advances in machine learning may soon allow AGVs to "think" their way around obstacles and handle new workflows.
Autonomous mobile robots (AMRs) are similar to AGVs in that they can safely transport inventory around a warehouse, but they also include advanced features that greatly expand the variety and complexity of the tasks they can perform.
The chief differentiator between AMRs and AGVs is that AMRs do not require pre-installed infrastructure to navigate through a crowded warehouse. Instead, they use an array of sensors to detect, map, and memorize the facility's features, using approaches like simultaneous localization and mapping (SLAM) technology. And they do it all while avoiding obstacles with real-time reflexes; communicating with other AMRs, warehouse management systems (WMS), and other software platforms; and even collaborating with human employees on picking and fulfillment tasks.
As one of the fastest-developing types of warehouse technology, AMRs seem to gain new capabilities every year. Many models can make their own map of an unfamiliar warehouse, then share that map with other robots, enabling companies to scale up their operations by simply rolling additional AMRs onto the floor, avoiding setup and installation hassles. Other models can use cloud-based software to optimize their path through the warehouse, detecting traffic jams or blockages and choosing new routes to get the job done faster.
The progenitor of this class was a squat orange robot developed by Kiva Systems to carry racks of goods to waiting human workers in what's known as a goods-to-person workflow. The technology worked so well that it was promptly taken off the market after Amazon.com purchased the company in 2012 and took it private.
But generations of new models have followed. Some have replicated the original Kiva design, while others have added new capabilities and attachments, such as spinning table-top belts for sliding parcels on and off conveyors, and tablet computers for communicating with human workers.
Other AMRs known as autonomous picking carts—or zone picking robots—operate on a robot-to-goods model. In that workflow, they automatically meet up with warehouse associates at specific racks or aisles, provide instructions on which goods to pick and which tote to place them in, and then whisk the completed order to another staffer at a packing station.
Autonomous inventory bots are essentially AMRs that count inventory on shelves instead of delivering goods to people or places. Outfitted with an array of sensors—computer vision, bar-code scanners, radio-frequency identification (RFID) readers, and more—they steer themselves around indoor facilities, constantly updating the DC's records on the quantity and location of goods. Inventory bots have also been deployed in retail and grocery stores, where inventory records are notoriously inaccurate.
Robotic picking arms have long been a familiar sight in industrial factories, where they perform precision tasks in automobile assembly or electronics manufacturing operations. But they're a relatively new entrant into the world of distribution operations, where they're starting to attract interest for their potential to boost fulfillment speed and accuracy.
Each unit includes a robotic arm with multiple joints and some type of "hand," known in the business as an "end-effector." The end-effector typically includes grasping fingers, suction cups, or some combination of the two, allowing it to seize objects ranging from the heavy (boxes on a pallet) to the light (garments packed in a plastic bag or small each-picks in a tote). Robotic arms rely on computer-vision sensors and artificial intelligence (AI) to help them recognize specific items and determine the best way to grasp them.
Flying drones and unmanned aerial vehicles (UAVs) are some of the latest entrants into the ranks of logistics robots, currently found mostly in pilot projects and innovation labs.
Like the birds in a forest, they vary greatly in size, speed, and capability. Some have whirring helicopter blades that allow them to hover, while other models look more like miniature propeller planes, able to soar on fixed wings.
Large warehouses might use drones indoors for inventory counting—a task they typically carry out by hovering over tall racks and scanning goods. But drones are also used out in the wider world, where they've been deployed to track trailers around truck yards, inspect infrastructure such as train tracks, and perform last-mile deliveries.
Choosing the best type of warehouse robot for your facility is a complex decision—considerations range from your budget and return on investment (ROI) goals to the condition of the building and your IT (information technology) infrastructure to labor availability in your particular market. But when you find the right solution, the benefits can be huge.
And if your perfect solution wasn't on this list, don't be discouraged. Chances are a robotics vendor somewhere is already working on a new design that will meet your needs. Recent technological advances have allowed developers to flex their creative muscle and respond to changing market conditions at almost lightning speed (think of the disinfecting bots that hit the market within weeks of the pandemic's arrival in the U.S., for example). It's safe to say robots have only just begun to find their niche in logistics.