Beyond the boxes: Mind-blowing insights from an Amazon Fulfillment Center tour

Takeaways from the Amazon Fulfillment Center Tour

Recently, Jason Hehman, our Industry 4.0 Vertical Lead at TXI, had the opportunity to tour an Amazon Advanced Robotics Sortable Fulfillment Center with the Manufacturing Leadership Council. He came up with three core innovation examples from which all industrial businesses can learn.

Computer Vision

Amazon’s Fulfillment Centers are leaders in using computer vision. They use the technology to ensure that every package that leaves their center passes a quality and fulfillment test.

Quality Assurance

Amazon uses computer vision to perform quality assurance checks to validate whether the human worker has selected the correct item to be shipped. When the shelf comes to the worker’s station (yes, in Amazon Fulfillment Centers, the shelves go to the workers; the workers don’t go to the shelves), the computer vision program takes a photo of the cubby where the ordered item is stored. Then, after the worker has retrieved the item, the computer vision program retakes a picture of the cubby. Then, the program compares the two images to ensure the worker picked the correct item.

Once the box is ready to leave the center, Amazon uses computer vision as a final quality assurance check to ensure the box size fits the expected size for the items ordered.

Automation

Computer vision was also used to automate part of the packaging process. For single-item orders, machines like the CW1000 utilize computer vision images to cut the cardboard precisely and fold it around the object. This kind of automation helps reduce material and shipping costs.

Computer vision technology has matured to the point where it is a highly effective tool for industrial operations. Its capabilities in quality assurance are particularly promising, as it can rapidly inspect products for defects, inconsistencies, or deviations from specifications with far greater accuracy and speed than human inspectors. Moreover, computer vision is now sufficiently advanced to handle basic robotic tasks such as cutting and folding, enabling increased automation and efficiency on the production line.

Edge Computing

Amazon does a brilliant job of utilizing edge computing for a company known for its cloud services.

In a sense, computer vision demands edge computing. The sheer number of high-resolution photos involved makes cloud computing a less attractive option. That’s why Amazon uses edge computing during the quality assurance process at the packer’s workstation. The local machine makes the validation onsite, saving data transfer time and resources.

Of course, Amazon still uses the AWS cloud for several tasks at Fulfillment Centers, including checking any order status updates right before the box leaves the facility.

Produced by NOVA Media