Inside Amazon's Smart Warehouse

Amazon is the world's largest retailer and its CEO, Jeff Bezos, is the richest man in the world for a very good reason: his company is better than anyone else at giving people what they want. Amazon quickly acquired its undisputed status as the heavyweight champion of the retail universe thanks in large part to its fast delivery times, the astonishing feat of transporting hundreds of millions of items, from guitar strings to pots and pans to auto parts, directly to your doorstep in 24 hours is nothing short of a modern logistical miracle, so how does Amazon do it? A super intelligent army. of slave robots for ingenious, if occasionally unscrupulous, management practices are also part of the answer and the modern voodoo of deep learning, all of which is realized in the most advanced storage rooms the world has ever seen, as well So join us today as we strap on our hubris jacket and journey inside Amazon's

smart

warehouse

s in the year to September 2020 with the global economy and the teeth of coronavirus and the bleakest job outlook in history.

Amazon reported global revenue of just under $350 billion, about double what it earned in 2017. By the way, just three short years earlier, not bad for a company that began operating in 1994. Amazon has built its empire on a platform of absolute unbeatable convenience for the end user: you, the customer, need a product, so you open the app or have a quiet word with Alexa and the next day, voila, it is at your doorstep. products from a to b quickly is not easy compared to other modern technology giants that barely need to exist in the real world, think Netflix, Google or Facebook, Amazon needs to shift a surprisingly large amount of high volume safely, accurately and Fast all day, every day, so how do you do it?

Last year, a senior Amazon executive described Amazon

warehouse

s quite poetically as a symphony of humans and machines working together, how does this symphony really work? Top, before you even log on to their website, Amazon has a pretty good idea of ​​what you're going to buy. This is all due to the semi-occult 21st century hocus-pocus that is deep learning, which Amazon has been harnessing to incredible effect. since around 2015. Simply put, an algorithm makes some assumptions about you based on your age, location, socioeconomic background, and purchase history. Then, hours, days or even weeks before you log in, you'll ensure your local warehouse is stocked with the right quantities of stock.

We're likely to consider buying this could be a certain bold swimsuit style that the algorithm anticipates will be the must-have new thing come spring. It could be the paperback novel for a viral TV adaptation soon to be published in January 2020, i.e. Amazon. The algorithm correctly anticipated the high incoming demands for face masks and we all know what happened there, so Amazon's

smart

warehouses, also known as fulfillment centers not to be confused with their post office-style sorting centers, very often They know what you want before you even do it or at least they know the moment you click buy now.

In the case of those late-night wine-inspired impulse purchases, once you've clicked our symphony begins in earnest, it's You may have read headlines in recent years suggesting that Amazon workers walk up to 12 miles per shift running between shelves, frantically picking up items, which is no longer entirely true. Amazon's modern fulfillment centers are largely patrolled by an army of squatting Roomba-like robots that pick up entire shelves, also known as pods, and bring them to a human picker situated at a stationary workstation. The story of this battalion began in 2012, when the company purchased the robotics company Kiva Systems, a market leader in warehouse automation, for the incredible sum of $775 million.

The company's flagship model, commonly known as the Kiva, measures about 30 centimeters tall and is capable of lifting 450 kilograms. of weight while traveling at about 3 miles per hour, substituting these chubby orange automatons for human workers makes a colossal difference to Amazon's bottom line. It has been estimated that Amazon warehouses can now hold 50 more stock and retrieve it three times faster. This reduces the total cost of fulfillment by about 40 cheaper, faster warehouses, which means more affordable products for the end user and, more importantly, products that are much more likely to be in the van driving around your street the next day.

Amazon is not about to stop. There, in a move that could be described as ruthless or inspired upon purchasing Kiva, Jeff Bezos changed the company's name to Amazon Robotics and told all of Kiva's former customers well-known names like Gap Walgreens and Staples that they would no longer be allowed. buy new products. The kiva technology, this of course, gives Amazon an untold competitive advantage since they implemented the kiva robot in their fulfillment centers in 2018, they already had 100,000, which has comfortably surpassed two hundred thousand. Amazon Robotics has been refining the design even further in the new iteration. Kiva known as Pegasus is 10 centimeters shorter, meaning it can stop further at the top, and uses half the parts, making it cheaper to make and maintain.

Amazon says the Pegasus can lift a hefty 600 kilograms and can be customized with a conveyor belt for sorting work. centers where Amazon reports that delivery errors have been cut in half thanks to Pegasus, naturally Amazon won't retreat there last summer; announced a newer, even slimmer robot called xanthus and it's coming to a fulfillment center near you, so how do you move these room bars without crashing into each other all the time the cloud-based software that operates what describe fairly how an air traffic control network run by AI coordinates the route of each robot this is all a matter of optimization which is the fastest route to reach a product that will not interfere with other robots in their own routes which is the optimal acceleration and deceleration speed Up to 800 robots can be deployed at a time on the warehouse floor although in practice the numbers tend to be kept low to avoid traffic jams when their batteries run out The robots are instructed to find the charging station Since robots took over the warehouses, changes have been implemented to improve their working conditions, for example, skylights are now covered so that the robot's sensors are not confused by the glare from the air conditioning units that blow down in some areas where humans work now blow sideways so as not to knock delicate items off the top of moving shelves to navigate a camera on the robot's landing gear reads QR codes embedded in the floor and individual sensors help the robot to slow down or swerve to avoid obstacles in its paths.

Compared to these hurried warehouse servants, some other robots working at Amazon seem almost monotonous and conventional. The so-called robo-stow robotic arm, for example, wouldn't look out of place in an old-school car factory, except that it can lift a hefty 1,200 kilograms and manipulate shipping pellets with an accuracy of a tenth of a millimeter, too. There are the labeling robots, nicknamed knock machines by their human co-workers for their incessant drumming noise, these can label up to one package per second and to get an idea of ​​what will happen in the coming years. Amazon recently purchased Canvas Technologies, a company specializing in autonomous robotic carts.

Imagine the most sci-fi beverage cart you can imagine. Robots, of course, are only part of the story within Amazon's smart warehouses. Even the most ardent company futurists admit that the notion of full automation is a decade away and even then it probably won't happen, so what about the human side? Amazon's management techniques in concept with all that automation have made the business surprisingly agile and bad by historical standards. In 2016, it was estimated that by bringing everything in-house as opposed to all the duplication inherent in a standard high street or mall, Amazon requires only half the employees of a traditional retail mine for every $10 million in sales.

What are all these humans, hundreds of thousands around the world, doing so well since Kiva and his robot? The airs have taken over, there is much less rush than before, but there are still many tasks that require skill and problem solving. The two most common functions still performed by humans are putting away and picking. When goods arrive at logistics centers, they are stowed by humans. on shelves or capsules for the robots to pick up later, the pickers then pick the specific item from the shelves as the robots pass by and then send it to be packed.

Picking side workers are encouraged to work fast to maintain their level. It is called a rate if the workers rate falls below expectations, employees can be disciplined and ultimately fired, according to a former employee, this rate can be a challenge to meet 120 items per hour when they started in the company and reach 280 items per hour just three years later. Mistakes are also punished according to the same former employees - workers were once allowed one mistake per 1,000 items, but are now only allowed one per 2,200. Rates only become more challenging during the main day, when Sales on Amazon skyrocket in a way that Amazon encourages.

For workers to set rates is through gamification or turning everything into a game, so instead of a plain old graph telling workers where their productivity is in relation to the rate, workers play and they compete in internal games with names like picks in space mission runner or castle maker, so essentially, the faster and more accurately employees choose actions, the faster their little pixelated car will move around the track, others Incentive schemes, such as internal currency loot, reward hard work with Amazon-branded products, such as bottles of water or tea. -shirts

amazon

has regularly found itself in the line of fire for its intense labor practices, with up to 14,000 serious injuries reportedly occurring at

amazon

sites in 2019, a rate per employee nearly double the industry standard, Deaths are rare but not unheard of in In the UK alone during the three years to 2018, ambulances were called to Amazon warehouses 600 times.

For its part, Amazon is keen to emphasize that it invests tens of millions of dollars in worker safety awareness programs, but it cannot be denied that injuries increase around the main day, as do ultra-efficient warehouses. ultimately a positive force, as Kovid 19 devastated large swathes of the mainstream retail landscape this year. Amazon has been processing up to 40 more orders than expected in the month leading up to March 23. Toilet paper sales alone increased 186 percent. of cough medicine soared by 862 and children's vitamins increased by 287 clearly in a world where going to the stores can be risky business amazon is filling a need and as such an increasing number of people depend on it if robots will steal our jobs so the outlook is unclear, but within Amazon it is clear that humans are still needed for many aspects of the job and even if robots one day can stock or pick as fast as humans, they face many crises, like leaking paint cans on a fast-moving conveyor belt. or identifying ripe bananas on site will continue to require the human touch for some time, whatever the future holds for Amazon's new $40 million robotics lab outside Boston, for example, or its tantalizing patent for fulfillment centers. floating blimp type, one thing can be guaranteed as long as We are all shopping Amazon will continue to deliver to you.