May 13, 2025
Image Courtesy of Amazon
Amazon’s Warehouse Robot Gains Human-Like Sense of Touch
Heralding the innovation as a “fundamental leap forward in robotics,” Amazon unveiled a robot, Vulcan, with a “sense of touch” that enables it to pick up and manipulate items without crushing them.
Vulcan is not the first Amazon robotics model that can pick up items. Older robots use a combination of suction cups and computer vision to move single items around.
However, in a press release, Aaron Parness, Amazon’s director of applied science, described the typical warehouse robot before Vulcan as “numb and dumb,” especially those working in commercial settings. He elaborated, “In the past, when industrial robots have unexpected contact, they either emergency stop or smash through that contact. They often don’t even know they have hit something because they cannot sense it.”
Vulcan’s biggest benefit is being able to stow items in the small fabric pods that Amazon uses to store inventory in fulfillment centers. The pod’s compartments are only about 1 foot square and contain up to 10 items, and older models didn’t have the finesse to remove one object without risking damage to others. Amazon said, “Fitting an item into or plucking one out of this crowded space has historically been challenging for robots that lack the natural dexterity of humans. Vulcan is our first robot with a similar kind of finesse.”
Already operational in warehouses in Spokane, Washington, and Hamburg, Germany, Vulcan is primarily being used to pick and stow items at the top and bottom of the 8-foot fabric stacks. That saves human workers from bending down or fetching ladders, which Amazon claims will improve worker safety and reduce injuries.
The AI-enabled robots also promise to make fewer mistakes than humans to support more “accurate” orders, avoiding inconveniencing customers and costly returns.
Vulcan can operate 20 hours a day with the ability to pick and stow approximately 75% of all the various items at Amazon’s fulfillment centers at speeds comparable to that of front-line warehouse employees. Trained on physical data including touch and force feedback, the AI-integrated robot “learns from its own failures” to support improving capabilities over time. Amazon noted that it does this by “figuring out how different objects behave when touched and steadily building up an understanding of the physical world, just like kids do. So, you can expect it to become smarter and more capable in the years to come.”
Amazon isn’t promising any cost savings from robots replacing humans, with the press release including quotes from warehouse workers on how Vulcan is making their jobs easier and safer as well as creating opportunities to help employees advance into robotics and other high-tech fields.
Parness told CNBC that Amazon fulfillment centers will still need human workers, especially for higher-tech roles that involve installing and maintaining the expanding robot fleet.
“I don’t believe in 100% automation,” he told the outlet. “If we had to get Vulcan to do 100% of the stows and picks, it would never happen. You would wait your entire life. Amazon understands this.”
BrainTrust
Jamie Tenser
Retail Tech Marketing Strategist | B2B Expert Storytelling™ Guru | President, VSN Media LLC
Neil Saunders
Managing Director, GlobalData
Brian Numainville
Principal, The Feedback Group
Recent Discussions
Discussion Questions
Does Vulcan appear to be a breakthrough in warehouse management and in optimizing online delivery?
Are fully automated warehouses the future, or will some human element always be necessary?
Poll
Loading …
This development is part of the path to the full automation of warehouses, which is key to making ecommerce more efficient and profitable. At present, most automation and robotics have limitations which means solutions are partial – but this is now starting to change. Amazon needs to be in the vanguard of this to maintain its leadership in the sector.
“Amazon isn’t promising any cost savings…” No it’s letting us do that; and how can we not(?): why else would anyone pursue this – or issue press releases – if it weren’t an end result. Some people will always be necessary…but the definition of “some” is “more than ‘none’.”
This reminds me of a saying related to AI. People ask if AI will replace humans. No, AI will replace humans who don’t use AI. It’s the same with factory automation. This is the future. Currently, it’s Amazon, but it’s a matter of time before others take advantage of warehouse automation. And just like AI, human involvement is still needed. Someone makes, oversees, quality checks, and more. And when the robot breaks, who fixes it? (Rhetorical question!)
Progress is good…but please keep them in the warehouse.
It’s the end of the world as we know it.
We’re starting to hear of in-store autonomous devices formerly used for inventory scanning only that have added arms used to manipulate items on store shelves, including some replenishment tasks. While early in the game, such capabilities seem likely to proliferate faster in controlled environments.
Robots that can glide about on their own and move stuff around will require more fail-safe engineering compared with those designed only for surveillance. They won’t come cheap, at least not at first, so I don’t think we’ll be overrun – this year.
The key feedback mechanism, as Amazon describes, is some kind of “touch sensitivity” that enables a robotic arm to learn the resilience, weight, grippy-ness/slippery-ness, and fragility of various items it handles. I expect the learning curve for that type of machine intelligence will be a matter of a few years at most.
These devices are likely to need a great deal of human maintenance and support – at least in the beginning. New jobs for a new era.
This is the path we are on. Just like agentic AI, where there may be less people doing a specific task, there will be someone overseeing these agents. Here, there will be more automation resulting in fewer people doing something but someone else providing a level of oversight of the robots.
More Discussions
How Important Is ‘Vibe’ in Physical Retail?
Not attempting to be too playfully groovy about this question: According to a recent report from…Will Pinterest Translate to Shoppable TV?
Will Pinterest lead the way in shoppable TV?How Can Retailers Improve Shoppers’ Sense of Closure When Concerns Emerge?
How can retailers better provide closure and proper resolution to today’s shoppers?
This development is part of the path to the full automation of warehouses, which is key to making ecommerce more efficient and profitable. At present, most automation and robotics have limitations which means solutions are partial – but this is now starting to change. Amazon needs to be in the vanguard of this to maintain its leadership in the sector.
“Amazon isn’t promising any cost savings…” No it’s letting us do that; and how can we not(?): why else would anyone pursue this – or issue press releases – if it weren’t an end result. Some people will always be necessary…but the definition of “some” is “more than ‘none’.”
This reminds me of a saying related to AI. People ask if AI will replace humans. No, AI will replace humans who don’t use AI. It’s the same with factory automation. This is the future. Currently, it’s Amazon, but it’s a matter of time before others take advantage of warehouse automation. And just like AI, human involvement is still needed. Someone makes, oversees, quality checks, and more. And when the robot breaks, who fixes it? (Rhetorical question!)
Progress is good…but please keep them in the warehouse.
It’s the end of the world as we know it.
We’re starting to hear of in-store autonomous devices formerly used for inventory scanning only that have added arms used to manipulate items on store shelves, including some replenishment tasks. While early in the game, such capabilities seem likely to proliferate faster in controlled environments.
Robots that can glide about on their own and move stuff around will require more fail-safe engineering compared with those designed only for surveillance. They won’t come cheap, at least not at first, so I don’t think we’ll be overrun – this year.
The key feedback mechanism, as Amazon describes, is some kind of “touch sensitivity” that enables a robotic arm to learn the resilience, weight, grippy-ness/slippery-ness, and fragility of various items it handles. I expect the learning curve for that type of machine intelligence will be a matter of a few years at most.
These devices are likely to need a great deal of human maintenance and support – at least in the beginning. New jobs for a new era.
This is the path we are on. Just like agentic AI, where there may be less people doing a specific task, there will be someone overseeing these agents. Here, there will be more automation resulting in fewer people doing something but someone else providing a level of oversight of the robots.