For consumers, the online purchase is fast, secure and easy. Learn how Amazon, always innovative, has mastered logistics and efficiency

As the holiday season approaches, it seems that everyone’s thoughts start turning to festivities with family, friends, entertaining, food, football, and, of course, shopping. Whether you celebrate the holidays or not, there are sales and deals to be had everywhere!

In recent years, online shopping has become a lucrative industry for online retailers, even retailers with brick and mortar store fronts have gotten into the lucrative, online market.

For consumers, the online purchase is fast, secure and easy. For online retailers, it is profitable. Take online retail giant Amazon, always innovative and known to masters of logistics and efficiency.

Amazon processes 35 orders a second… 2,100 orders a minute… 126,000 orders an hour… 3,024,000 a day which adds up to 1,103,760,000 online orders a year!

This is for an average day. Consider a more mind-boggling Amazon statistic, on Amazon’s 20th anniversary—AKA Prime Day—they sold 400 items per second!

Additionally, Amazon found that every 100 milliseconds of latency costs them 1% in sales. For these guys, latency matters. A lot.

Given that the retail industry is based on profitability, one can understand Amazon’s requirement for operational excellence at all times.

Amazon is an Optimized, Efficient, Well-Oiled Machine

Amazon’s process is broken down into steps. From the time items arrive into the warehouse, to the point where these are sold and delivered to consumers, every step is a choreographed, systematic approach, balanced between persons and robots.

The distribution centers are enormous. For example, the warehouse in Phoenix, Arizona is 1.2 million square feet, which is the equivalent of 28 football fields. There are more than 15 million items stored here in endless rows of seemingly jumbled consumer goods (which is actually a method called random stow).

The ‘random stow’ method appears to allow inventory to be stored haphazardly, but this storage approach is actually based on a complex computer algorithm. It results in shelves housing completely unrelated items, for example, one shelf contains a Monopoly game, a roll of duct tape and bottles of dial liquid soap, all tucked in next to one another.

Part Man, Part Robot

Amazon believes that they have optimized the ratio of humans and machines in its warehouses worldwide and utilizes a seemingly perfect blend of a robot-human assembly line. They believe that the efficiency lies with the number of robots and humans that they use.

Amazon introduced Kiva robots into its warehouses in 2014. Their purpose is to automate the whole picking and packing processes at large warehouses. They started with 15,000 in 2014 and as of 2015’s third quarter, Amazon has doubled up to 30,000 Kiva robots.

The robots are in use to do the heavy lifting and to identify and carry items. People sort products into place in the storage cages or stay at single stations, packing and organizing.

Amazon emphasizes that the robots aren’t meant to replace people, rather, that the robots compliment people and make them more efficient.

In fact, as of November 2014, Amazon had not yet eliminated any jobs with the implementation of the Kiva robots, rather they hired more people. Amazon won’t divulge just how many jobs were added, but overall, the company has hired 61,110 employees since 2011 (the year before Amazon bought Kiva—now Amazon Robotics). This has roughly doubled Amazon’s employee base in the last few years!

If you aren’t familiar with the Kiva robot, it is a square-shaped yellow robot that moves on wheels, is about 16 inches tall and weighs 320 pounds. They travel 5 miles per hour and lift packages weighing up to 700 pounds.

How does Amazon process 35 orders a second?

Amazon incorporates automation technology in multiple, massive facilities—in April 2015 there were 96 reported warehouses in the US alone, but that number is speculated to be bigger—and can still deliver in two days.

Much of the process is automated—once the buy button is clicked and the user has gone through the purchase flow, associates are notified via hand-held scanners that tell associates exactly where to retrieve the product. More, the scanner then uses an algorithm to determine the next optimal item to be picked.

Items are placed in a big box, called a tote, and filled boxes are sent to the ‘SLAM’ line on a conveyor belt. SLAM stands for ‘Shipping, Labeling and Manifesting.’ This automated process also includes a quality control checks confirming that the weight of packages are correct—if they are not, packages are set aside for associates to review.

Packages are then ready and are sent down shoots designated for the geographical location they are headed for. At this point, it becomes a human process. Associates fill boxes with the goal of having no wasted space. Think of it as a live game of Tetris.

Automation Technology Landscape

Amazon isn’t the only e-commerce warehouse operator adding robots to increase their efficiencies. Corporations like Coca Cola, Walmart, Nike, Ikea, FedEx, and UPS are looking to reduce fixed racking and conveyor costs in favor of dynamic storage and mobility methods. And it’s not just the Kiva robot, there are literally hundreds of suppliers of warehousing automation technology: Westfalia, Egemin, Schaefer, Swisslog, Vanderlande, Daifuku, Intelligrated, JBT, Knapp and Thiele to name just a few.

The Bottom Line: Latency Matters

Automation technology relies on efficiency and process. It’s not even a matter of long downtime, even low latency matters—in fact it costs organizations (and not just e-commerce) dearly:

Ensuring Efficiency

e-Commerce is the fastest growing segment of retail. In 2013 it represented less than 10% of overall retail, but has been growing at nearly 15% annually and is expected to continue at that rate for the remainder of the decade. Online retailers are having to change their operations to meet this growing demand. [i]

It goes without saying that achieving operational excellence isn’t just with the distribution teams, rather, it goes deeper. The technology itself requires expert technical maintenance as inefficiencies will cost big dollars.

Training to mitigate latency issues requires more than just a one-time course. It can be a matter of continuous practice to keep skills sharp. Utilizing simulation software keeps overhead costs minimal in comparison to costly downtime.


Tell us, What do you do to mitigate latency?

Share This