The Eye of AI #7

The Eye of AI #7

Computer vision is a truly transformational technology for the retail industry, but it has applications that touch nearly every aspect of our lives. This is our weekly summary of what happens when cameras, computers and great ideas get together.

NASA's Self-Flying Aircraft Project

A box-of-tricks fastened to a helicopter can help it learn to fly itself. Photo: NASA

A full list of Future Sources of Frustration that None of us are yet Prepared For must surely include sitting in a traffic jam in an autonomous car, especially if the whole kerfuffle includes autonomous horn honking. Well, the obvious way to avoid these kinds of problems is to hop into a driverless helicopter and soar over the melee. This kind of 3D-thinking has led a team of NASA boffins to develop AIRVUE, a cutting-edge pod designed to enhance the computer vision systems of electric air taxis and drones. Mounted beneath a helicopter, AIRVUE collects vital environmental data, such as weather conditions and obstacles, to create a comprehensive dataset that will be shared with AAM manufacturers.

A Paper Over the Cracks

Visual inspection of concrete is an unreliable test of its safety. Photo Photo by

Concrete has been used as a building material for thousands of years in part due to its malleability, excellent thermal mass properties and fire resistance. It is the second most used material in the world after water, yet we still struggle to know how safe concrete buildings are. A recent paper from researchers in Chengdu, China aims to crack the code, with a machine vision system that has been trained to identify fractures while ignoring the noise inherent in rough surfaces.

How do you See in the Dark?

We are each of us rolling a ball of dung, but some of us are looking at the stars. Photo: University of South Australia

Way back in our second edition we explained how the eye structure of the preying mantis was helping improve depth perception in computer vision systems. Well now, engineers are taking a page from nocturnal dung beetles, those tiny poop-rolling navigators of the night, to help drones and robots find their way in the dark. Inspired by the beetles’ trick of using the Milky Way as a cosmic GPS, an Australian team has developed an AI sensor that can spot the Milky Way even when things get shaky. Unlike individual stars that blur out of sight with motion, the Milky Way stays rock steady, making it the perfect guide for low-light navigation. So, thanks to some beetle-inspired brilliance, we might soon see drones cruising the night sky, guided by a galaxy far, far away!

Improving the Airport Experience

Simple improvements can make a huge difference to travel experience. Photo: Air India

Air India just upped the ante in travel tech with the launch of ‘AEYE Vision’ on their mobile app, becoming the first airline in India to use AI-powered computer vision for real-time trip updates. This feature lets users skip tedious data entry—just point your phone camera at your boarding pass, baggage tag, or ticket, and voila! The app instantly fetches all your trip details. Whether you need to check flight status, track your bags, or add trips, AEYE Vision has your back. And the best part? Air India plans to roll out even more cool features soon, like image-based destination search and AR-powered travel details. Travel just got a whole lot smarter!

Fast Reliable Soil Analysis

Soil structure is paramount in plant health. Photo: Photo by

Agritech research is currently driving a lot of new application in computer vision, so it's not a surprise that a new study takes a high-tech approach to analysing soil structure. Using 12 different machine learning models to quickly and accurately classify six types of soil based on texture, some models hit over 99% accuracy. The Bayes Net model was the top performer with a 99.83% accuracy rate, proving that computer vision can make soil analysis faster, easier, and more precise. This tech could revolutionize soil management by making it easier to decide what crops to plant and how to manage fields for the best yields.


That's everything for this week. Please keep an eye on the SAI Group blog for everything that we're thinking and talking about, including a new article this week about innovations in self-checkout security.

Got some cool tech to share? Whether its your own project, or just something you saw and thought "I want people to know about this!", let us know about it and we'll include it in upcoming editions.


要查看或添加评论,请登录

SAI的更多文章

社区洞察

其他会员也浏览了