Delivering the Future, which has recently been lead by Amazon for the past few years, showcased the latest technologies behind its operations. In Seattle, this year's update included updates to the drugstore goods and drone delivery capabilities.

A major theme of the event held this year in Nashville included updates on AI within shopping experiences as well as the firm's use of computer vision, adding further trim to delivery times for packages. While no new robotics systems have been touted, the conference did shed some interesting insight into how the company integrates its existing offerings.

Just a short time after the event, we sat down with Amazon Robotics chief technologist Tye Brady. The conversation has been an annual event for the last few years, giving us a chance to dig deeper into how Amazon's robotics story has changed in the past 12 months and how the next 12 will evolve.

According to internal figures from Amazon, the company currently deploys 750,000+ robots in its U.S. fulfillment centers. It's the same public number the company proclaimed in 2023. But it's not all of the news, however. The number counts only 750,000 autonomous mobile robots, or AMRs.

These are the wheeled systems that the company calls, which have been patrolling the floors in the Amazon warehouses since the firm bought Kiva Systems way back in 2012. The firm's tote robots it calls drive-train systems come in numerous different models, including its latest, an autonomous system known as Proteus that it first unveiled at Re:Mars in 2022.

AMRs make up the majority of Amazon's fleet but not the whole of it. Other form factors carve out their own space on the floor. The second big class of robots is that of the Amazon robotic arms, with Robin, Cardinal, and Sparrow all sorting and stacking items into their own classes.

Newest to the Amazon Robotics family is Sequoia, unveiled at the 2023 Delivering the Future event. The name might garner from the redwoods of Northern California and possibly reflect the dimensions and scale of the system. Sequoia is an automated storage and retrieval system that is similar in concept to those from companies like AutoStore.

The first Sequoia system rolled out in 2023 at a Houston-area fulfillment center. On Wednesday, Amazon said a system 5x its size now sits at the heart of a massive warehouse in Shreveport, Louisiana. The fulfillment center itself isn't new, but it is being dramatically expanded to sprawl across more than 3 million square feet.

Amazon is clearly big and rich enough to design entirely new greenfield robotic fulfillment centers from the ground up. It is actually concentrating on retrofitting brownfield warehouses that already exist. That's less expensive, but it also means that the company must work in a manner to accommodate existing operations of "fix the airplane while it's flying" in Brady's words.

The Shreveport center will be one of Amazon's "Gen 12" buildings, ultimately deploying 10 times as many robots as their forebears — a number the company has yet to reveal. But in addition to those robots come more robot-centric jobs: 25 percent more RME (reliability maintenance engineering) roles than before, Brady said.

The company points out that when the sites reach full capacity in Shreveport's 55 football fields of fulfillment operations, it will have employed 2,500 humans. Automation enthusiasts will tell you these technologies allow humans to focus on the things they can do that robots can't. Brady is a big proponent of the notion.

Asked what jobs humans are still better suited for, he answers, "Problem solving, common sense, thinking with reason, understanding the big picture, understanding the context. Some physical tasks, as well."

While Digit stole the show at Agility's 2023 appearance, Amazon was mum on the humanoid front. It's true the company has been testing what bipedal robots can do within its fulfillment centers, including a pilot with Agility it announced last year. But since completing the pilot, things have gone quiet.

"We're still learning," Brady says of the Agility partnership. "It's slow and steady. 'R&D' is the best way I can capture it." The slowness, he explains, is a product of finding ways such technology could slot into existing workflows.

We start with the problem we are trying to solve," Brady says. "When you have a piece of technology and say, 'Hey, how do I apply this?' that's a dangerous path when you try to force things. The fact is, in our fulfillment centers, we have a lot of nice, poured concrete floors. Wheels are pretty good. But we also have stairs and uneven terrain when we start to go outside.".

Brady confirmed that the partnership is still active but hasn't had much more to say to TechCrunch.

One such partnership that has emerged strongly in the last few months is that of the UC Berkeley spinoff Covariant. As one might remember, the startup's founders Pieter Abbeel, Peter Chen, and Rocky Duan had joined the company in August along with around 25% of its employees. This also marks the effort towards scaling foundational models in an industrial setting.

For instance, Amazon says its robot arm, Sparrow, "can now handle over 200 million unique products of all different shapes, sizes, and weights." But there will always be edge cases. What's decisive now will be how humans and better-trained AI systems handle these moving forwards. Covariant will play a key role in that, operating in those kinds of massive datasets to fine-tune things like product pick and placement.

"We're up and going and starting to work on some really meaty, very applied problems for machine learning," Brady says of the deal.