Published in a deluge The most notable news at this week’s Delivering the Future event in Seattle was that Amazon will begin testing Agility’s Digit, a move that could bring bipedal robots to fulfillment centers across the country. It was information. These are still small steps, and early-stage deals like this don’t mean anything big in the future.
Take Agility’s Ford pilot, for example. At the time, the startup was exploring last-mile delivery as a possible future. Shortly thereafter, the company began focusing his Digit production solely on warehouse and factory operations.
Last April, Amazon named Agility one of the first five recipients of its $1 billion Industrial Innovation Fund. Joining the fund doesn’t guarantee Amazon will use your technology in the future, but it’s clear that the retail giant is – at least – interested in the possibility.
“The real purpose of the Innovation Fund is to explore what’s possible in the world,” Ty Brady, chief engineer at Amazon Robotics, told me in an interview this week. “It’s also about understanding practical examples.”
The executive added that while Amazon Robotics has so far only worked on wheeled transportation, legs have a lot of potential.
“We’re interested in walking robots,” Brady says. “I think the ability to move across different terrain is very interesting. We’re also interested in what works and, frankly, what doesn’t. It’s really interesting. I don’t know if that’s a good thing or a bad thing. We’re experimentalists at heart. We’ll figure it out. We’ll do a pilot and see if it’s a good thing. We’re going to see how it works.”
The company’s focus on wheeled AMRs (autonomous mobile robots) dates back to its acquisition of Kiva Systems in 2012. Kiva Systems’ platform forms the foundation for Amazon Robotics as a whole. There are currently 750,000 AMRs deployed across the company’s warehouse network. The company has also launched non-AMR systems, including picking arms like the Sparrow, which was launched during the same event last year.
It’s difficult to overstate how profound an impact Amazon’s efforts have had on other industrial robotics fields. First, the company has increased the pressure on competitors to automate to meet rising customer expectations for same-day and next-day delivery. Another is that his decision to stop supporting Kiva customers outside of the Amazon ecosystem directly led to the creation of industry giants like Locus Robotics and 6 River Systems.
For Amazon to integrate the system into its growing robotics ecosystem, the system must demonstrate productivity gains. It’s not about innovation for the sake of innovation, it’s about identifying all the advantages of delivering products to customers in less time. Including drones.
It remains to be seen how humanoid robots specifically, and bipedal robots more generally, will be integrated into ruins. Another major hurdle is that the new system must fit the almost inconceivable size of the company.
Many startups are currently vying for the humanoid robot crown, including 1X, Figure, and Tesla. Agility’s Digit is the least human of them all, but it has a lot of money and a huge head start. The company recently opened a new factory in Salem, Oregon, and claims it can produce up to 100,000 digits a year when fully online.
There’s a lot of excitement about this category, but proving things on a large scale is another matter entirely. Whether Digit succeeds or fails in its prescribed tasks could have a significant impact on the trajectory of humanoid robots in general. Just as Kiva Systems proved to be a major catalyst for AMR, if Amazon succeeds in deploying Digit at scale, suddenly everyone will want to get their hands on humanoid workers. Probably.
The biggest conversation around form factors is the fact that humans build workspaces for other humans. This includes ledge height, terrain, aisle width, and the bane of his ARM’s existence: stairs. From this perspective, humanoid robots suddenly make a lot of sense. The reality is that most companies operate from brownfield sites. This means their warehouses and factories are typically not built with specific automation solutions in mind. Humanoid robots can successfully enter brownfield sites.
Of course, Amazon has the resources to build any facility it wants, so it’s no surprise that many of its robots are efficiently operating at greenfield sites. These limitations are less of a concern for Amazon than for other competitors, but it would certainly be ideal if an effective system could be integrated into existing workflows with minimal friction.
Image credits: Amazon
However, Brady acknowledged that Digit is not the final word in Amazon’s plans for mobile operations.
“Once you start holding it, [sensing, compute and actuation] When you put them together in interesting combinations, really unique things start to happen,” he says. “When it comes to mobile robots, we are world leaders. And now we are working hard on manipulating not only packages but also objects. And by integrating them, we see all the possibilities. I’m really looking forward to it.”
That might mean entering in another way. Amazon, for example, knows how to build both AMRs and robotic arms. If one can effectively put the latter on top of the former, they will have some kind of mobile operation.
“If you look at an agility robot, you can think of it as a mobile manipulator,” Brady says. “That’s interesting to us. We haven’t done much research on bipedal robots, so we’re particularly interested in locomotives. But we definitely want to use them as identification systems, manipulation systems, sorting systems. We do everything we can to innovate for our customers and improve safety for our employees. We’re getting there with our core fundamentals .”
If for some reason Digit fails to land, that doesn’t mean the end for it or for bipedal robots in general. Maybe it just doesn’t fit comfortably into Amazon’s existing workflow. Maybe the robot isn’t quite up to the scale of Amazon, or maybe Amazon hasn’t reached the stage yet where his Digit makes sense.
In any case, it would be wise for anyone with even the slightest interest in bipedal robots to pay attention to this. The pilot could have a huge impact on how we think about this category going forward.