I am a volunteer in this network. Getting on a big UAS page will accelerate use of new bridge inspection technology like this. If you fly on the other side of the bridge you are going to lose signal as the main path, pylons, frame elements block the pilot's view. A hobby pilot can fly a drone. It's easy. Someone operating in the role as lead field tech could monitor the data and keep communication open to larger control structure through ICS, ATC, whatever.
There will be an inability to inspect all drones without the use of submersibles and drones. There are bridges that are partially or totally inaccessible after disasters. BVLOS is going to collapse the level of risk for infrastructure inspection. Up close, cheap, no bodies in the air. Or in the water.
Differential thrust, redundant sensors allow for precise movement, but really large metal objects are a pain in the butt. They can cause loss of transmission. If you could reference the size of bolts or rivets on a bridge, you can get positional control from that . The same technology used to sync up a Pokemon player with a "shrine" uses something like a histogram to obtain a feature or class of features to obtain energize a relay to a motor or another field device. Using three drones in an area is like using repeaters for WiFi. Waves. A UAS can capture all the waves. I like all the waves.
My challenge is to evolve an automation process and advance standards in that task. That is really exciting to me! PA is full of bridges. Very hilly, and I live in a valley area along a river. Cheap inspection makes bridges cheaper, no better. IR sensors can train up to 3 mm, but at that distance you risk losing control, signal. Also, put a camera on top of the drone, a little Runcam or something.