A bird detection system such as this (less so species detection and more so avian detection) would be very useful at airports where bird strikes are quite common, and where humans are regularly used to sight for birds near/around runway approaches.
I’m surprised to see they're not marketing towards the aviation (no pun intended) industry too.
The identification part seems pretty slick. I'm curious about the "curtailment" part - if a bird is heading toward the chopper, what can you do? Slamming on the brakes on 100m+ blades seems... intense.
Wow, this is way more interesting than I was hoping for when I clicked on this link. Identifying birds computationally is an enormously tricky problem, but because they're tasked with reducing wind turbine's killing
protected species their focus is almost exclusively on raptors, which is an much easier nut to crack. Even if you can't always separate a highly protect species (ex: Golden Eagle) from a similar looking, less protected species (Ex. Turkey Vulture) 100% of the time you should be able err on the side of caution and reduce mortality events.
Randall Munroe published the 'Tasks' (Park or Bird) XKCD comic almost 6 years ago, talking about the task of identifying photos of birds being a 5 year research problem.
Readit News
I’m surprised to see they're not marketing towards the aviation (no pun intended) industry too.
Presumably with a bit of surveying pre-construction you can make a good guess at your expected curtailment time and bake that into the financial model
Tennis?
Not a terrible prediction.
Dead Comment