While robots taking over the world is an idea that has gone viral in recent years thanks to videos like this one where a robot declares they will destroy all humans, we may not have reached the Jetsons level we all hoped they would by now. Warehouse robotics in the logistics industry isn’t very common — Currently, around 80% of warehouses are manually operated, according to Robotics.org.
However, recent strides in robotics innovation has lead to more and more warehouses beginning to use robots in the warehouse management sector. But what do we mean when we talk about robots? What are the benefits and drawbacks of warehouse robotics, and can they be used in small to medium sized businesses?
What is the Current Technology?
The current robot technology is a lot different than what has gone viral today. Unlike robots like Sophia, which are built to resemble and express human tendencies, industrial robots are humanoid, but built for function rather than form. Supply chain logistics is a complicated process and difficult to automate, especially due to the specificity of each warehouse’s layout and the need for varied, specialized tasks to fulfill each order. Additionally, warehouses are messy — the constant hazard of obstacles such as pallets or boxes in the aisles makes simple automation of even the most menial tasks difficult. Still, some companies have managed to solve for the problem in a variety of ways.
Amazon, for example, began installing its Kiva robots into warehouses in 2014 to automate the picking and packing processes. These tiny robots have cut operating costs by over 20% alone for Amazon. This saves the company around 22 million for each fulfillment center. This small square machine weighing at 320 pounds can haul packages over double its weight across the warehouse and have cut cycle time from around 75 minutes to just around 15 minutes, according to Business Insider. Amazon’s warehouse robotics fleet helps increase productivity. They work by bringing entire shelves to human pickers, who then pick and pack the items for shipping. Amazon has thus solved for one problem that can take hours and up to 8-10 miles of walking per employee in even the most organized warehouse. But what about the whole process? Can robots be used to pick, pack, and ship all the orders themselves?
Can a Warehouse be Fully Automated?
The short answer is no, not yet. Robots like the Boston Dynamics Atlas Robot can avoid obstacles and lift objects, which could help theoretically in the packing and putaway processes; however, the logistics of training a robot like Atlas per warehouse is nearly impossible when weighed against the costs. While CEO Elon Musk dreams of having a fully automated Tesla factory, his Gigafactory currently only reaches halfway. Robots like the Self-Navigating Autonomous Indoor Vehicles are able to navigate the factory floor freely without guidance using a digital map. Their main purpose is to move materials through the factory. Another is a robotic arm which can handle over 350 kg. Still, Musk requires humans to run and work alongside his warehouse robotics crew in his advanced factory.
Is Warehouse Robotics Feasible?
Clearly, use of robots in the warehouse is on the rise for large enterprise businesses. However, every warehouse is unique, and the infrastructure required to train the robots for the many different sized rows, shelves, boxes and pallets would cost a lot. The robot AI would need to be trained to learn the environment, and would need specific parameters imputed every time. Additionally, the types of robots used within a fully automated warehouse would have to be compatible. For example, an AI enabled forklift would have to be able to call a humanoid robot like Atlas or Kia to its location in order to shift the items without error. You would also need a WMS system to organize and manage the robots throughout the process in order to streamline picking and packing and centralize the process.
Another issue with a fully automated warehouse involves the act of picking and packing the items themselves. Currently, many different items are jumbled into one section for picking due to an efficiency algorithm. While humans can select and pick the correct item, robots cannot currently identify objects without a marker. This would add to the costs of infrastructure. Additionally, humans are able to use judgement when picking up fragile items. A robot, on the other hand, would have a tougher time handling objects with the right amount of force.
However, some of these issues could be remedied over time, should the cost of these solutions be reasonable. For example, robots with built-in scanners could easily identify the products they need using barcodes. Suction lifting for fragile-marked objects could reduce the chances of the object breaking while being picked. Additionally, humanoid robots could eventually replace forklifts, should they be able to both handle heavy loads and reach high up on the shelves in a warehouse.
What About SMB’s?
Currently, the cost is much too high for robots to feasibly be used in SMB warehouses. The cost is not only due to the robot, but also time spent training the AI. Another issue is that you must equip your warehouse to be robot friendly. This could mean doing anything from installing tracking on the floor for guidance to tagging every product, shelf, door and station to ensure the robot knows where to go and what to do at all times. However, as AI personal assistants get cheaper and easier to acquire thanks to products like Amazon’s Echo Dot and Google Home, we could be seeing the advent of cheaper, easier to implement robotic logistics technology in the future.