The company’s technology takes street level imagery from companies such as TomTom and uses new statistical and image processing techniques to accurately identify assets and tag them with unique GPS coordinates. The neural network detects objects of interest visible in the images, weeds out the duplicates that can appear in multiple photographs, and ascertain detailed measurements such as height of signs or poles, width, and distance from verge.
The system could also be used to audit speed limit signs, for example, and subsequently identify which ones need to be changed in the event of a nationwide altering of speed limits, or to reduce street clutter where multiple poles have been installed for different signs.