OH NO! You have a project due tomorrow. Quickly rushing to your room, you open your computer and head to Amazon.com to place an emergency order. After a quick browse, you find your desired item and press buy. Your order is being processed within seconds of confirmation, and within hours, a perfectly folded cardboard box makes its way onto your front porch alongside all your desired project materials. But how does your click translate to a package on your doorstep in just a few hours?
Companies are increasingly turning to hardware and software automation to address the rising demand for faster delivery times, labor shortages, rising wages, and surging order volumes. Across its 175 fulfillment centers around America, retail conglomerate Amazon uses over 750,000 robots to perform highly repetitive tasks, freeing employees to deliver to customers more efficiently. The result? A network service so efficient that over half of all Amazon Prime orders arrive within just two days, according to TechCrunch.
Veritas traveled to an Amazon fulfillment center located in Tracy to learn about the story behind each package on the front porch. For most small items purchased on Amazon, this journey starts in these fulfillment centers, placed on pallets of randomly assorted goods. Robots transport these pallets across a sleek concrete floor embedded with fiducial-tracking tags — markers similar to QR codes — that help robots pinpoint their locations.
By using technology and computer vision, Amazon items do not have to be sorted onto their storage shelves. They can be placed on any pallet in the warehouse, and computer databases will automatically keep track of their location. When an order is placed, robots go to the nearest pallet with that item and transport it to an employee for processing.
After being processed by an employee, items move onto a conveyor belt and are grouped by order into crates to be sent to packing stations. There, cameras figure out optimal box shapes to handle items, and workers pack items together for shipment. Then, giant hydraulic arms, which possess a 360-degree range of motion, laser sensors, and a 165kg loading capacity, sort these packages. Once sorted by delivery location, workers load packages onto delivery trucks, and they’re ready to be delivered to your doorstep.
Similarly, other delivery companies, such as FedEx, are embracing automation to optimize delivery times and package handling with robotics. Late last year, FedEx partnered with Redwood City-based robotics company DexterityAI to launch DexR, a new robotic arm that autonomously loads packages into vans, optimizing storage in delivery trucks.
Using computer vision, DexR scans packages to identify their shapes and dimensions and then uses artificial intelligence to map out the best possible arrangements to fit them into a truck. Think of it as a three-dimensional jigsaw puzzle, where each puzzle piece is its own specially shaped box.
Avinash Verma, head of supply chain and operations at DexterityAI, emphasized the company’s current priorities for management and organization.
“Our focus right now is the logistics sector, so we work with companies like FedEx and UPS,” Verma said. “We are working on things that involve loading a truck or moving boxes around in warehouses. We are doing palletizing and depalletizing with real-time robotics. … With a lot of the algorithms we have, we can work with hundreds of thousands of different objects, and computer vision is like our eyes: It’s just a tool.”
When merchandise and other deliverables are stored in ever-growing warehouses, keeping track of every item can be difficult. With companies like UPS delivering over 5.7 billion packages yearly, better logistics systems are integral to keeping up with ever-increasing delivery demands.
“Everyone now just swipes their phone and orders on Amazon,” Verma said. “This isn’t only in the US. This is worldwide. There is a lot of work that needs to happen in warehouses … and there are not enough people to do that manual work.”
With many industry giants in the logistics sector, such as Amazon, facing labor shortages and high turnover rates, automation has become an even more lucrative business in the last few years.
“Manual work was done by people, and there was a shortage of workers,” Verma said. “Because they were all manual operations, there wasn’t a lot of information about the operation.”
With all these developments in the delivery industry, Verma is optimistically steadfast about the future of robotics. As warehouses become more advanced and demand scales up, robotics companies will face challenges keeping up, something Verma firmly believes DexterityAI has the capacity and technology to do so.
“I think the exciting part is that in the future, these robots are going to get better,” Verma said. “The positive part of machine learning and AI right now is that the more it gets advanced, the better it gets. … We are at a transformational point in time where a lot of … tasks are going to get automated.
