Amazon is testing new technology that will help employees at delivery stations more efficiently identify and sort packages before they’re loaded onto delivery vans.
It’s called Vision Assisted Sort Station (VASS) and it uses computer vision and projection technology similar to the Vision-Assisted Package Retrieval (VAPR) system in Amazon’s delivery vans.
When packages arrive at Amazon delivery stations from local fulfillment and sortation centers, employees sort and stow them into bags which then get loaded into vans by their delivery route to reach their final destination—customers’ doorsteps. During this process, VASS will create a static buffer area, holding multiple packages, where it spotlights packages with visual cues to help employees quickly identify the right ones without the need to look at a screen or device. In parallel, the destination bags are brought to them, so employees no longer have to walk. This dramatically improves efficiency and simplicity through the sorting process.
"VASS is a powerful example of how we're innovating at Amazon to make work easier for our teams," said Chris Harris, Amazon’s director of logistics engineering. "VASS helps simplify decision-making for our employees and speeds up the process of getting packages where they need to go.”
This innovation builds on VAPR, an Amazon technology embedded into delivery vans that automatically project a green “O” on all packages that need to be delivered once the van arrives at a delivery location, and red shapes on all other packages. Through an audio and visual cue, VAPR prompts the driver to confirm it has found the right packages before the driver even needs to enter the cargo area.
The technology, which can locate and decipher multiple barcodes in real time, was built by training machine learning models to recognize different labels and packages in various lighting conditions and package characteristics. It also removes the need for drivers to use a mobile device throughout the process.
Based on early VAPR tests, Amazon teams saw a 67% reduction in perceived physical and mental effort for drivers and more than 30 minutes saved per route.
Real benefits, real results
The technology’s cameras, being explored for an under the roof application in our delivery stations, scan packages on a conveyor belt so when they arrive at a stow station, employees instantly know their destinations.
VASS shines a green O, indicating to the employee which package needs to be picked (and a red shape on those that shouldn’t be picked) and placed in the corresponding destination bag, which is brought directly to the employee for better ergonomics.
“It's sophisticated technology under the hood,” Harris added, “but what our employees experience is something simple, intuitive, and genuinely helpful."
Early testing of VASS at Amazon’s Innovation Center DNZ3, in Dortmund, Germany, showed promising results:
- Reduced mental load for employees, who no longer need to memorize destinations or constantly reference their handheld devices or screen.
- Increased accuracy with a substantial decrease in missorts in pilot stations.
- Greater efficiency through improved sort time per package: Employees were able to manage a buffer of up to 10 packages simultaneously, versus just two with conventional methods.
- Enhanced flexibility: The system decouples sorting performance from machine sequencing, allowing employees to work at their own pace.
Looking ahead
Currently in pilot phase, VASS is scheduled for implementation across some of Amazon’s delivery stations in Europe and the U.S. starting in 2027.
Amazon's last-mile delivery technology team is already working on enhancements that will incorporate more advanced computer vision algorithms. VASS, as an iteration of VAPR, is an example Amazon’s ability to innovate and scale across its operations, and shows our commitment to employees and partners, to keep delivering smiles to customers.
