An Amazon warehouse is a flurry of activity. Workers jog around a man-made cavern plopping items into yellow and black crates. Towering hydraulic arms lift heavy boxes toward the rafters. And an army of stubby orange robots slide along the floor like giant, sentient hockey pucks, piled high with towers of consumer gratification ranging from bestsellers to kitchenware.

Those are Kiva robots, once the marvel of warehouses everywhere. Amazon whipped out its wallet and threw down $775 million to purchase these robot legions in 2012. The acquisition effectively gave Jeff Bezos, its 52-year-old chief executive, command of an entire industry.

He decided to use the robots for Amazon and Amazon alone, ending the sale of Kiva’s products to warehouse operators and retailers that had come to rely on them. As contracts expired, they had to find other options to keep up with an ever-increasing consumer need for speed. The only problem was that there were no other options. Kiva was pretty much it.

Starting to make inroads

It’s taken four years, but a handful of start-ups are finally ready to replace Kiva and equip the world’s warehouses with new robotics. Amazon’s Kiva bots proved this kind of automation is more efficient than an all-human workforce. The new robots being rolled out look different, partly because the industry is still experimenting and partly because of patent issues. Some focus on picking items off shelves, others zoom around with touch screens. All are aimed at saving retailers money as they race to get their wares to your doorstep as quickly as possible.

Amazon has about 30,000 Kiva robots scuttling about its warehouses across the globe. Dave Clark, the retailer’s senior vice president of worldwide operations and customer service, estimated the addition of the bots reduced operating expenses by about 20 percent. According to an analysis by Deutsche Bank, adding them to one new warehouse saves $22 million in fulfillment expenses. Bringing the Kivas to the 100 or so distribution centers that still haven’t implemented the tech would save Amazon a further $2.5 billion.

“To be great in e-commerce, you need to be sophisticated inside the warehouse,” said Karl Siebrecht, CEO at Flexe, which bills itself as the Airbnb for warehouse space. Amazon was the first company to confront the challenge of picking a virtually endless variety of goods from warehouse shelves and combining them in a single box for home delivery. Now that e-commerce is a growing part of the retail trade, more companies are paying attention.

But it’s really only Amazon that has this kind of technology at scale, thanks in large part to Kiva. The world’s biggest retailers, including Wal-Mart, Macy’s, and Target, have yet to populate their warehouses with widespread robotic systems. They rely on the old method — otherwise known as humans: Hordes of pickers and packers who send boxes down conveyor belts.

For the new breed of robot makers, the potential market is wide open. Logistics companies that run their own warehouses started designing automatons while ambitious engineers saw the hole Bezos blew in the market and jumped in. One start-up was even founded by former Kiva workers. The race to automate was on.

The modern warehouse is a rectangular box with 40-foot-high ceilings, loading docks on both sides, and often, thousands of parking spaces for staff that swells during peak shopping seasons. Lately retailers have been demanding something new: floors that approach the Platonic ideal of flatness, a feature that makes life easier for the technologists managing fleets of warehouse robots.

While automation has long been the looming threat to industrial workers, there’s reason to think their situation is about to become worse. There were 856,000 warehouse workers in May, according to the Bureau of Labor Statistics. The average wage for those workers is about $12 an hour, said David Egan, head of industrial research for the Americas at commercial real estate firm CBRE Group. Minimum wage hikes being considered and enacted nationwide could drive labor costs higher, especially in locations close to city centers — sites that are in high demand as retailers chase Amazon into the realm of same-day delivery.

Investing in automated distribution centers may offer companies a hedge against such uncertainty, though it will come at the expense of an increasing number of Americans who rely on warehouse work to survive. Not only could robots pare labor costs over the long term, they may protect employers against labor shortages, a particularly scary proposition for the biggest retailers come Christmastime.

Robots can help improve speed and accuracy and increase productivity per square foot of warehouse space at a time when the growth of e-commerce is driving up commercial rents.

All these imperatives have led most big warehouse users to experiment with new types of automation. “Some companies are implementing in a big way, and the majority are doing at least a pilot project on one small area or in one warehouse building,” said Raj Kumar, a partner at consulting firm AT Kearney. That includes Wal-Mart, which has been using robots to ship apparel from its e-commerce website, he said. Wal-Mart, which didn’t respond to an e-mail requesting comment, has also been experimenting with flying drones that photograph warehouse shelves as part of an effort to reduce the time it takes to catalog inventory.

“Warehouses are very high-tech places,” said Bruce Welty, co-founder and chairman of Locus Robotics, a firm that’s developed bots to work alongside, rather than replace, human workers. “Because the only way you can take costs out is automation.”

No joke

Locus is a spinoff of a company called Quiet Logistics, which owns two warehouses in Massachusetts, a gateway of sorts for e-commerce goods being distributed across America’s northeast corridor. Welty and his co-founders based their distribution business on Kiva’s robots. They built software systems around the Kiva bots to improve efficiency as they moved goods for such retailers as Zara, Gilt, and Bonobos. Then Amazon blew it all up.

“I said once to my board — casually, almost as a joke, ‘Boy, we’d really be screwed if Amazon bought this company,’ ” said Welty. “But I never really thought this would happen.”

Such a move was unprecedented in warehousing: the mass withdrawal of one specific type of technology.

Usually, one company would buy another and keep selling the tech to all the usual customers. That’s where the money was, after all. Not Amazon. It wanted the Kivas all for itself.

Amazon declined to comment.