Amazon's New AI Technology VAPR for Electric Vans and Package Retrieval

TapTechNews October 9th news, Amazon just released an announcement to explain the new AI technology on its freight cars. It is said that Amazon will deploy 1000 Rivian electric vans equipped with Vision Assisted Package Retrieval (VAPR) function in early 2025.

It is introduced that Amazon began to introduce electric delivery vans in the summer of 2022, and currently more than 15,000 are in operation across the United States. TapTechNews noticed that Amazon also announced that it will bring its custom vans to Europe, including a fleet of thousands of electric vans.

Amazon said that with the new AI solution of Vision Assisted Package Retrieval (VAPR), couriers will no longer need to spend a lot of time organizing packages according to the delivery point, reading labels or manually checking key information such as the recipient's name and address to ensure that they are delivered to the right person. Now they just need to grab the package and go, and then VAPR will show a green light on the corresponding package at the right place (other packages are red lights), so as to find the right package in a few seconds.

Thanks to audio and visual cues, VAPR will give cues before entering the delivery area to help the courier find the right package. In addition, this technology also avoids the need for the courier to use mobile devices throughout the process.

BloomfieldLogistics courier BobbyGarcia said, Previously, I might need 2 to 5 minutes to empty a shelf pallet and organize packages for the next delivery point. Now with VAPR, the whole step only takes about a minute, which makes my life much easier.

It is introduced that VAPR uses Amazon's AR-ID machine recognition technology, which is a computer vision technology originally developed for distribution centers and can automatically identify items during storage or picking stages, thus avoiding the process of manual barcode scanning by staff.

VAPR can identify different labels with different lighting conditions and different package characteristics through machine learning models, and can locate and interpret multiple barcodes in real time.

Moreover, this technology is also optimized for the in-vehicle environment and can be used in conju nction with specially designed automotive light projectors and cameras to achieve integration with the delivery route navigation system of the van. Amazon said that VAPR is applying a variety of AWS technologies to practice, including the machine learning platform SageMaker and IoTGreengrass.

According to Amazon's early tests, the physical and mental consumption perceived by its delivery staff has been reduced by 67%, and an average of more than 30 minutes of delivery time can be saved per route. If we consider Amazon's more than 390,000 DSP couriers, more than 100,000 delivery vans, and the delivery of millions of packages every day, VAPR can save a large amount of energy and time for its staff.