Amazon is using a new AI model to help determine the most efficient type of packaging for each item, in an effort to speed up the process of meeting its sustainability goals.
In combination with other packaging innovations at the company, the Package Decision Engine model has seen Amazon avoid two million tonnes of unnecessary packaging worldwide since 2015.
The reduction has been achieved by reducing the number of cardboard boxes, envelopes, paper dunnage, tape and paper bags used to send purchases to Amazon customers.
Built on the Amazon Web Services (AWS) cloud, the model can predict when a more durable product like a blanket doesn’t need protective packaging, or when a fragile set of items like dinner plates might need a sturdier box.
The AI model uses deep machine learning to keep abreast of Amazon’s evolving packaging options, while natural language processing helps it make decisions for certain key words like “grocery”, “screen” or “stoneware” to spotlight suitable options, and computer vision helps it assess the products dimensions and spot defects.
Subscribe to Sustainability Beat for free
Sign up here to get the latest sustainability news sent straight to your inbox everyday
It marks a step away from Amazon’s previous process of using physical testing to help optimise the packaging, with Amazon saying that the machine helps to automate this process and free up time for other sustainability work.
Already in use across fulfilment centres in the UK, Europe and North America, the retailer is set to expand it for use worldwide – which will involve exposing it to new languages, and the different packaging types used in different companies.
“We wanted the ability to quickly identify the most efficient packaging option for each item, while also predicting how safely each product would ship,” said Amazon packaging innovation team senior manager Kayla Fenton.
She continued: “The use of AI through the Package Decision Engine has allowed us to advance our packaging efficiency work at scale quickly, and it has worked so well that we’re implementing this technology across Amazon’s broader global footprint.”
