Today we announce a world first: the ability to find regions of rust within an image.
Seems easy enough, but to thoroughly examine 1,000,000 images daily or identify points of interest for a drone on the fly? That’s a bit harder. Our algorithm can efficiently process images both onboard and offline.
Give it a try: SpotRust running on Amazon Web Services
Finding and evaluating rust has always been a manual process. Numerous patents exist for finding and evaluating rust but they only work if you can attach electrodes or spray chemical indicators. For $100,000 you can have a corrosion engineer drive out to your site (or fly to your ship) and evaluate structural elements like pipes, tanks, mechanical assemblies and other equipment for problems related to rust.
Or now, for less than 1/10th as much, you can use our algorithm to filter points of interest with photos and video collected on site, summarize the findings and then deliver the results to your expert remote consultant.
Consultants, why travel if you don’t need to? Evaluate images from multiple projects spread across the world. Or if you do go on-site, run our algorithm on your helmet cam video to find any random corners you may have missed.
In the growing world of network-controlled industrial exploration, visually inspecting rust is a killer-app. Proactively identifying spots where corrosion management coatings peeled or cracked can save a full replacement years later during plant turnaround.
Radically extend the lifetime of your capital assets by leveraging a combination of regular quadrotor survey flights, thorough quantitative image analysis, and timely spot-maintenance.