Amazon’s Vulcan robot brings tactile sensing to warehouse automation

Amazon‘s new Vulcan warehouse robot represents a significant advancement in tactile robotics, enabling machines to physically interact with objects in ways previously challenging for automation. By incorporating touch sensing and machine learning, this system allows robots to rummage through shelves and identify products, potentially addressing a major limitation in warehouse automation while establishing a new approach to human-robot collaboration in fulfillment centers.

The big picture: Amazon has developed a touch-enabled warehouse robot that can search through shelves to locate specific products for customer orders, marking a meaningful advancement in robotic dexterity.

• The robot, named Vulcan, uses tactile sensing to manipulate items on shelves and identify products, capabilities that have historically been difficult for robots to master.
• Already operational in facilities in Hamburg, Germany and Spokane, Washington, Vulcan works alongside human pickers on the same fulfillment lines.

How it works: Vulcan combines a conventional robotic arm with specialized appendages and touch sensors that allow it to detect and manipulate objects in crowded storage spaces.

• The system features a spatula-like tool for poking into shelves and a suction device for grabbing items, with sensors on multiple joints that detect edges and contours of products.
• Machine learning algorithms interpret the sensor signals and control how the robot interacts with objects, enabling it to push items around and identify specific products.

What they’re saying: “When you’re trying to stow [or pick] items in one of these pods, you can’t really do that task without making contact with the other items,” explains Aaron Parness, Amazon’s director of robotics AI who led Vulcan’s development.

• Ken Goldberg, a roboticist at the University of California, Berkeley, notes that “Amazon stores many different products in bins, so rummaging is necessary to pull out a specific object to fill an order.”
• Parness emphasizes Amazon’s collaborative approach to automation: “We don’t really believe in 100 percent automation, or lights out fulfillment. We can get to 75 percent and have robots working alongside our employees, and the sum would be greater [than either working alone].”

Why this matters: The development addresses a persistent challenge in warehouse automation by enhancing robots’ ability to handle varied products in complex storage environments.

• Vulcan aims to spare human workers from physically demanding tasks by handling items on shelves that are positioned too high or too low.
• While a significant step forward, experts note that robotic tactile abilities still have substantial room for improvement before matching human capabilities.