More than Meets the Eye: How AI-Enabled Vision is Helping Deliver Your Mail

computer vision delivers mail

It turns out seeing is actually very hard. In the U.S. alone, over 40% of the population needs glasses; by 2050, studies predict that nearly half of the global population will be nearsighted. Fortunately for humanity, technology is helping us navigate a world beyond 20/20 vision. Through computer vision, we’re teaching computers to see and interpret surroundings in a way that is transforming how we do business and live our lives. 

What is computer vision?

Vision AI, also known as computer vision is a science that deals with how computers use digital images and videos to gain a higher-level understanding of our world. Using deep learning models, machines can better identify and classify what they “see” in images and react accordingly.  

If this concept sounds familiar, that’s because it is! Back in the 1970s, computer vision used optical character recognition (OCR) to interpret handwritten texts for the blind. Today, computer vision helps industries in automotive, healthcare, eCommerce, and of course, government agencies such as the postal office. Some benefits of vision AI include: 

  • Boosting productivity: Computer vision systems can carry out repetitive tasks quicker, simplifying people’s work.

  • Producing better products: Good computer vision systems commit fewer mistakes, resulting in faster delivery of quality products.

  • Increasing Cost-effectiveness: Computer vision leaves less room for faulty products, saving companies money used to fix flawed processes.

Computer vision in postal automation

In mid-December 2020, the United States Postal Service delivered as little as 62% of first-class mail on time — the lowest level in years. Their target is around 96%.   

After a complicated year, USPS unveiled a promising 10-year plan that will invest billions into its infrastructure and workforce. The government agency recently deployed computer vision systems in 195 locations that, according to reports, will cut the time it takes to track a missing package from several days to less than 2 hours.

While the USPS has used OCR to scan mail and packages for a long time, the new OCR that manages USPS’ imaging workflow will now live as a deep learning model inside the new technology—making it a gamechanger for postal automation.

What’s most exciting about the level of innovation happening inside the US postal service (aside from knowing we’ll successfully get our mail) is that it opens up a new world of possibilities for other government agencies and even other industries.  

Vision systems across industries

Today, we can see fascinating uses of vision systems across industries. Here are some examples of how technology continues to change the way we experience daily life:


Google sees over ten billion searches related to skin, hair, and nail conditions per year. The tech giant’s new teledermatology app, Derm Assist, combines computer vision and machine learning to help users detect skin conditions from their smartphones.

Sports tracking 

Vision systems went from being virtually unknown to highly prolific in the last five years. With technology like precision ball-tracking and real-time player capabilities, computer vision is changing the way we experience live sports.

Inventory control  

Last year, Walmart tested a robot with cameras that perused stores making sure shelves were stocked. 

You may have used computer vision on your phone recently. If you’ve used the camera feature on the Google Translate App to read a foreign-language menu, parked in a ‘smart garage’ that lets you know what spots are open, or used your iPhone camera as a tape measure, then you’ve used a vision system, too. 


As vision systems technology evolves, the potential for new applications will grow. On the machine side, component developments now offer a greater variety of cameras that can be used to create new lenses, intricate robotics, and more. On the computer vision side, faster processors, deep learning, cloud computing, and data integration software present us with opportunities beyond what we could have imagined.

New technologies are emerging and improving at lightning speed. At 10Pearls, we’re here to help your enterprise both adopt modern technologies like computer vision and stay ahead of what’s next for your industry. Talk to us.