You'd think so.

My neighbor's wife decided to go on a walk in the middle of the night. She has dementia. Last year, coldest night of the year (-20f).

Once called in, there were helicopters using infra-red imaging and dozens of people looking for her. There were dogs, there were police, there were friends, and there were firefighters.

She was found four hours later in the back of a car. She climbed into someone's car and fell asleep. The car was about 1000 or so feet away from their house, but down a driveway and out-of-sight for the most part.

She was okay, other than being very cold. She's still doing okay, but it shows that it's not super easy to find someone.

I wonder if there is any sort of protocol for this or do people make it up as they go along.

I can imagine that a spur of the moment search like that would be chaotic at the outset.

Seems like a great space for a smartphone app that could define a grid direct searchers and use GPS to track their movements.

Yeah seems like some kind of tracking device is a requirement for people with advanced dementia. Have had a couple of friends whose grandparents have disappeared; one of them was found dead the other one...is still being looked for. Is there any reason we can't use gps tags of some sort, other than privacy concerns?

You can buy safety pendants that contain a GPS tracker hooked up to the cell network. It has an emergency button and can be located through the service's website. Some even have a speaker and mic so you can talk to an agent through it. I'm sure they could add a feature where if the pendant is a certain distance from home then have an alert go out to a list of contacts. Buy two of them for grandma if you really want to make sure that she is never not wearing one.

Apple's HealthKit/ResearchKit was extended (maybe 2 years ago?!) to include geo-fencing for basically that use case. Will try to dig up the article or github repo in a bit...

Suppose that, on average, it takes you one second to ascertain that there definitely is no body on 10m² of ground.

Then it would take 100,000 seconds to search a km². Rounding down by 15% or so, that's a full 24 hours.

A 1.7 mile radius circle has an area of 20 km² (again rounding down). So, by this extremely simple model, searching that circle would take two man-months at eight hours of searching a day, seven days a week. At ten times that speed, it still would take a week or so. Most people would get bored and move farther out before that week was over.

So, is that 10m² per person-second a good guess? I won't know, but looking around my living room, I doubt I could do double that rate there, so I (very, very tentatively) conclude it isn't completely bollocks.

There will be places where searching can be done much faster, but also ones where it is much slower, if only because navigating there is slow.

>A 1.7 mile radius circle...two man-months at eight hours of searching a day, seven days a week

They had 200 people that first week plus it's really an arc of a circle in the direction they set out and on one side of the river only.

>Most people would get bored and move farther out before that week was over.

That could be the problem there, that and people not searching in a systematic way. The referenced stories all seem to point to people all being found relatively close to where they set out. Maybe that needs to be pointed out to searchers. Also a grid kept of what has been searched and what not so the same ground isn't revisited over and over.

Also as was pointed out, the local sheriff was in charge and had basically no expertise in this.

Their plan for the run and the terrain would constrain the highest probability area quite a bit more than a circle around the ranch.

Yes, an arc of even 60 degrees looks too big

1.7 miles is a very long distance when you are lost and wandering.

True.

Assuming a flat trail and traveling in a straight line, that would be nearly 40 minutes at a comfortable walk.