Mostly fear. There are a lot of people suffering out there right now and the risk of that leading to a war is real. This is a salient reminder that militaries can send a drone anywhere and mow people down.
The forestry industry is nice and all, but tech like this is going to literally kill people. That will be the most interesting application in this decade.
- What if adversaries develop full AI-tech with weapons mounted on them? How can a nation defend against such a threat?
- What if democracies around the world get wiped out by such a tech which has no public opposition in communist/ authoritarian states? They will and are developing such technology.
- We just saw the use of (non-AI) drone military tech in the Armenia-Azerbaijan war. Being defenseless is not an option.
If it wasn't the CCP, it would be Russia again, or if not Russia, then another emerging power. In 10 or 20 years it could be India, or the fragmented parts of European Union that combine to form authoritarian regimes if the swings toward autocracy continue.
Every military force is investing in drone technology because any military force in the world is investing in drone technology, and while the US was the leader for a long time, the push to outsource manufacturing and globalization led to a broad distribution of the skills, tools, and manufacturing base to build modern drones.
China is no more the root cause of the arms race than Thog was when he hit Ugh with a stick some time in ancient history. Its humans and tribalism all the way down.
You're not considering the historical consequences of your perspective. And I made no such claim that China is the root cause of the arms race, that's a strawman.
Historical counterfactual: What would've happened in WW2 in Asia if the US had no aircraft carriers because some pacifists convinced the administration that investing in arms was just tit for tat tribalism and contributing to an arms race?
If you can't explain why the US should not invest in drones today but they should've invested in aircraft carriers in the 1930s and 1940s, then you're holding a sack filled with little more than misplaced confidence.
This is done a lot, you are not used to the idea of 'Swarm/Formation' so you are using it to cure everything.
One long range drone with thermal cameras (and other wavelengths) is all you need. No formation is needed or wanted.
I still rarely see even one drone used. We need to start with one, if its useful.
I can give you many locations with invasive species, swarms won't solve that. Unless you give the swarms guns.
But one drone with a gun would kill pests fine. I think even with hunting humans most killing wouldn't need formations.
That said SAR already use drones to help search when weather permits.
For autonomous systems one thing I find interesting is how far they have to go in this sort of domain versus a good human pilot for example: https://youtu.be/bFVvUeP_oqY?t=36
Perhaps carrying one of these folded up ready to deploy in an emergency wouldn't be a terrible idea as the availability of SAR imaging goes up. Much cheaper to carry that than some of the GPS/beacon devices sold to hikers. Image detection is already a common ML tool for Earth observation analysis so building in something that looks for that marker in a known area of forest sounds possible.
https://www.westmarine.com/WestAdvisor/Selecting-a-Radar-Ref...
That said if you're doing anything serious away from people carrying a GPS and EPIRB/SPOT is way more effective. The latter working even in the most remote locations. It's no good being able to signal if no one is even looking for you.
Others are working on drone charging systems that are faster than the unit's flight time: https://dronelife.com/2020/08/01/a-drone-battery-that-charge...
To the best of my knowledge, all SAR drones operate above the tree canopy. The forest I live by has, even at peak occupancy, more people sized animals than people. It makes thermal imaging through a dense canopy difficult at best. A good friend spent 2 years as a coast guard pilot stationed near me, and he spent the majority of his time doing SAR training. He was the one to describe the issues with thermal imaging above the canopy. We discussed using drones at length, and his main issue was the risk that they might randomly enter his flight path and damage a rotor. Operating below the canopy would eliminate this risk.
https://phys.org/news/2019-07-iceland-trees-razed-vikings.ht...
Keep in mind that timber cruising is an industry, so there are also plenty of focus on software and hardware to speed up inventories: https://www.fs.fed.us/forestmanagement/products/measurement/...
Sorry, not sorry
A friend did his PhD in lidar mapping of trees and forests, with a focus on identifying risk trees. He primarily focused on roadside trees that were a danger to drivers, but that was in part due to the ease of lidar mapping from a vehicle. He did trials with drone imaging, but they had to be manually controlled, which made them no better than people.
I live near a national park, and they average several lost hikers and roughly one death a year. The majority of those are due to poor planning and injury (FFS always check in at a ranger station, tell someone your schedule/route, and stay on the trial if you are injured!). A system like this with thermal cameras could find injured people much faster than a search party every could.
Half the battle of fighting a forest fire is understanding its location and boundaries, and tracking/predicting its movement. A system like this could be augmented with smoke particulate sensors and thermal cameras.
Many ecosystems are being devastated by invasive species. Wild hogs, anacondas, airborne pests. Early attempts to track 'murder' hornets were hampered because the scientists could not follow their transponders effectively.