Men rescued from Pacific island after writing SOS in sand
1–10 of 123 posts
Re: Men rescued from Pacific island after writing SOS in sand
#2Good news. Thanks for that!
Re: Men rescued from Pacific island after writing SOS in sand
#3How often is the best earth-facing satellite imagery updated?
It'd be really cool to sic handwriting recognition on it to continuously look for "SOS".
Re: Men rescued from Pacific island after writing SOS in sand
#4Call this a freely given and viable but hard to monetize idea - this is a very tractable area for software and drones. Same with locating people in lost wilderness areas. You have a programmatic way to organize coverage areas, you have access to above-human audio and video capabilities, and it reduces the resources/time ratio that is critical in rescues.
This was the tragic and avoidable story that convinced me this was possible - https://www.nytimes.com/2016/05/27/us/missing-hiker-geraldin...
There was a limited geographical footprint and coordinated, modestly sized fleet of quadcopters with mics and/or thermal imaging could probably cover 20 square miles/day. Same couple be applied with optical/higher altitude drones for people lost at sea. You could supplement sensors with basic face/voice recognition in less remote settings. You could supplement with something like Planet in more remote/ocean settings.
Does something like that exist already?
Re: Men rescued from Pacific island after writing SOS in sand
#5Call this a freely given and viable but hard to monetize idea - this is a very tractable area for software and drones. Same with locating people in lost wilderness areas. You have a programmatic way to organize coverage areas, you have access to above-human audio and video capabilities, and it reduces the resources/time ratio that is critical in rescues. This was the tragic and avoidable story that convinced me this…
Well, there are just personal locator beacons which you can rent and tell rescue personnel exactly where you are if you need a rescue.
Re: Men rescued from Pacific island after writing SOS in sand
#6I think the days of being stranded on islands are over. But what about oceans and being lost at sea? Would the film "Cast Away" make sense with today's satellite and object detection technology?
Re: Men rescued from Pacific island after writing SOS in sand
#7Call this a freely given and viable but hard to monetize idea - this is a very tractable area for software and drones. Same with locating people in lost wilderness areas. You have a programmatic way to organize coverage areas, you have access to above-human audio and video capabilities, and it reduces the resources/time ratio that is critical in rescues. This was the tragic and avoidable story that convinced me this…
If you were lost at sea / in the wilderness, would you trust your safety and eventual return to a machine learning model picking up your body heat? I know I'd want actual people in the aircraft.
Re: Men rescued from Pacific island after writing SOS in sand
#8Call this a freely given and viable but hard to monetize idea - this is a very tractable area for software and drones. Same with locating people in lost wilderness areas. You have a programmatic way to organize coverage areas, you have access to above-human audio and video capabilities, and it reduces the resources/time ratio that is critical in rescues. This was the tragic and avoidable story that convinced me this…
If you were lost at sea / in the wilderness, would you trust your safety and eventual return to a machine learning model picking up your body heat? I know I'd want actual people in the aircraft.
You can do both. Predator/Reaper/Global Hawks [1] [2] working alongside human aircraft, deployed from land or aircraft carrier. If you don't trust the ML, have humans in the Vegas desert watching the feed. I imagine military sensor platforms are very good for identification of humans.
> Five Reapers can provide 36 hours of combined surveillance coverage in Afghanistan with individual sorties lasting up to 16 hours; a further five vehicles increases this to 72 hours
For reference Afghanistan is roughly 251k square miles in size. Ten Reapers to cover 251k square miles for 72 hours at a time. Caveat, I have not done the math what the O&M costs are per hour for such a drone fleet.
Alternatively (or in concert?), Planet Labs supports 50cm resolution, and you can buy their imagery with a credit card. They are imaging almost the entire Earth daily (although they might exclude oceans? I imagine for the right price, they'll retask or capture a grid on demand). [3] [4] It's not heat sensing, but there might be signal to be found.
[1] https://en.wikipedia.org/wiki/General_Atomics_MQ-9_Reaper
[2] https://www.pacaf.af.mil/News/Article-Display/Article/220396...
[3] https://www.planet.com/why-planet/
[4] https://www.planet.com/pulse/more-spectral-bands-50cm-global...
Re: Men rescued from Pacific island after writing SOS in sand
#9Call this a freely given and viable but hard to monetize idea - this is a very tractable area for software and drones. Same with locating people in lost wilderness areas. You have a programmatic way to organize coverage areas, you have access to above-human audio and video capabilities, and it reduces the resources/time ratio that is critical in rescues. This was the tragic and avoidable story that convinced me this…
Or... a dude in a helicopter with FLIR? Sometimes, low tech is best, honestly.
Of course they already have that. The problem isn't coverage usually, it's density of the terrain features and vegetation, and an unknown search radius in some cases.
That, and unprepared people with no way to signal a search vehicle if one comes near. That's a human problem though, not a tech problem.
Re: Men rescued from Pacific island after writing SOS in sand
#10Call this a freely given and viable but hard to monetize idea - this is a very tractable area for software and drones. Same with locating people in lost wilderness areas. You have a programmatic way to organize coverage areas, you have access to above-human audio and video capabilities, and it reduces the resources/time ratio that is critical in rescues. This was the tragic and avoidable story that convinced me this…
If you were lost at sea / in the wilderness, would you trust your safety and eventual return to a machine learning model picking up your body heat? I know I'd want actual people in the aircraft.
I have been involved in lower grade search efforts. This is not either/or, this is supplemental. If you've never been in a scenario, particular with children, time is of the essence particular with temperature or terrain hazards. Time is the enemy, and the goal of this would be reduce time to cover a search area. If you find it interesting, https://www.outsideonline.com/2059616/how-backcountry-search... is a good read.