Innovation

AI tech in Amazon vans to help drivers identify packages faster

Machine learning models have been trained to recognize various labels and package types.

By early 2025, Amazon plans to deploy 1,000 electric delivery vans equipped with Vision-Assisted Package Retrieval (VAPR), an AI-powered solution designed to streamline package retrieval for drivers.

Although each stop only requires a few minutes for drivers to search for packages in the van, the cumulative time adds up significantly when delivering to over 100 customers daily.

VAPR eliminates the need for drivers to organize packages by stop, read labels, or manually check key identifiers like customer names or addresses. Instead, drivers can simply look for VAPR’s green light, grab the package, and move on.

The concept of VAPR originated in early 2020 when a team at Amazon Transportation envisioned a future where technology seamlessly assists drivers. “We had to consider factors unique to delivery, like lighting and space constraints inside vans,” said John Colucci, Product Manager at Amazon Transportation. Thanks to feedback from drivers during testing, VAPR is set to be integrated into 1,000 electric vans from Rivian by 2025.

VAPR utilizes Amazon Robotics Identification (AR-ID), a form of computer vision first developed for Amazon’s fulfillment centers, which identifies items without manual barcode scanning. AR-ID can process multiple barcodes simultaneously, even in challenging lighting conditions.

Machine learning models have been trained to recognize various labels and package types, making VAPR highly efficient in assisting drivers.

  • Press release