I was talking about this very thing recently- apparently there are very few engineers with the experience to make decisions about certain major rail bridges- just two worldwide in one example. Experience cracks in such a bridge in UK and you have to fly one in from Japan- time and money wasted if all you have to do is stream data from a drone. If you need more detail add video/photo data plus intuition from a local engineer's tablet or smartphone.
The real value-add comes when you ingest the rich metadata into a machine learning analytics platform like IDOL from Hewlett packard Enterprise. Combined with the Vertica Big Data Analytics Platform and you have cost-effective means of managing and insuring assets.
Look at 360Globalnet for practical examples of utilising streamed data in insurance market.
On the property/casualty side, we’re looking more and more closely at drone technology, the underwriting side you can survey properties much more efficiently, much more accurately, and you can survey more dimensions than a human could on their own. Temperatures, and structural elements, as opposed to just what the naked eye can see. On the claims side you can assess damage in certain territories for where disasters have occurred, and which can immediately result in not just more accurate information but faster claims payout which is what we want for all of our insureds.
http://www.insurancejournal.com/news/west/2016/06/09/411464.htm