Videos and computer vision for infrastructure management

Where are our assets? What is the condition? and how well does the infrastructure perform to fill operational purpose? How has the operational purpose changed during the lifespan of the structure? These are the most important questions that are normally presented while considering the domain of infrastructure asset management. Vionice is presenting at Väylät & Liikenne -seminar on this topic on Thursday 8th September. This blog post is a summary of the article we produced to the seminar magazine. Hope you get here the core idea and if interested to get more detailed info find your way to get the whole article. Of course you may write me and ask for it ;)

Why to use mobile phones for street-level videos ?

The concept generated by Vionice Oy is practically a simple use case of computer vision, sensor fusion and machine learning. The low cost sensors like mobile phones are able to be used as sensors. Naturally the requirements set by the use case and customer are to be filled with the most suitable hardware. But what makes the smartphones so attractive to use? While considering a scalable solution, that would be as cost effective as possible, it is good to start with simple and scalable solutions. Phones are perfect platform to scale for broad operational usage. Smart mobile devices also fit well for demonstrations in those cases where there is estimated to be a need for more solid installations that require investments. What ever is the camera installation, the video feed from the field is the first and most essential part of the system. On our experience video is more usable if compared to photos because they produce more data and make the information production more easy. Compare it if you would try to drive a car and close your eyes to open them once every second. There would be left out a lot of information that your brain would not be able to register.

Revolution in asset management

The raw-data is the core question for the whole process to automate information production. Computer vision is a tool to datamine knowledge from the frames. The possibility to automate manual work makes it possible from cost perspective to make it in reasonable price. Infrastructure asset management is going to be revolutionized because of the productivity leap in automated data production - well come on-board, fasten your seat belts.