Sponsored Link
Email This Post Email This Post

A Physicist Helps Amazon Reduce Package Weight

Amazon
A Physicist Helps Amazon Reduce Package Weight

Reducing the packaging of an order can save in materials and in shipping costs, and at Amazon’s massive scale, saving even a penny or two per package adds up to big bucks.

But it appears the retail giant is going well beyond pennies. It has used sophisticated technology that has reduced per-shipment packaging weight by 36% over the past six years. It has eliminated over a million tons of packaging that it says is the equivalent to over 2 billion shipping boxes.

Amazon enlists the help of a physicist to help it reduce packaging. Matthew Bales heads up machine learning within Amazon’s Customer Packaging Experience team.

In a blog post published this week, he said when he started at Amazon in 2017, it was doing a lot of physical testing of products, “but not a scalable mechanism that could assess hundreds of millions of products to identify the optimal packaging type for each product.”

“Statistical tests were the first piece, but they are essentially only useful when products have already been shipped in more than one package type. We wanted the capability to predict how a product would fare in a less-protective, lighter, and more sustainable package type. And once you’re in that predictive space, you need machine learning,” Bales explained.

“Machine learning” approaches, particularly deep learning, thrive on big data and massive scale, and a pioneering combination of natural language processing and computer vision is enabling Amazon to hone in on using the right amount of packaging, the company said.

IBM describes deep learning as “attempts to mimic the human brain—albeit far from matching its ability—enabling systems to cluster data and make predictions with incredible accuracy.”

Amazon uses customer feedback in training the technology. One interesting challenge for Amazon: how did the vendor package the product before sending it to one of its fulfillment centers. “For example, a ceramic mug may be packaged in clear plastic bag, or in a sturdy box.”

Amazon’s team knew that product images on the Amazon site weren’t helpful when selecting packaging. “For example, a multipack of LED bulbs might be illustrated by a picture of a single, unpacked bulb, suggesting it is fragile, yet the multipack is, in fact, safely packaged by the vendor and doesn’t require additional packaging. It is best shipped in its own container.”

So the team used Amazon’s own image data: “When products are delivered to fulfillment centers, many are sent via conveyor belt through special computer-vision tunnels equipped with cameras that capture images of the products from multiple angles. These tunnels are used for many things, including ascertaining product dimensions and spotting defects.”

The blog post described the challenges in more detail. Bale called the success of deep learning in reducing packaging a triple win: “Reduced waste, increased customer satisfaction, and lower costs.”

Ina Steiner on EmailIna Steiner on LinkedinIna Steiner on Twitter
Ina Steiner
Ina Steiner
Ina Steiner is co-founder and Editor of EcommerceBytes and has been reporting on ecommerce since 1999. She's a widely cited authority on marketplace selling and is author of "Turn eBay Data Into Dollars" (McGraw-Hill 2006). Her blog was featured in the book, "Blogging Heroes" (Wiley 2008). Follow her on Twitter at @ecommercebytes and send news tips to ina@ecommercebytes.com. See disclosure at EcommerceBytes.com/disclosure/.

Leave a Reply