← Back to the blog

The first hour: what 200,000 search-and-rescue cases tell us about being lost

6 June 2026 · 10 min read · By the FreeBusy team

In the United Kingdom, someone is reported missing every 90 seconds.

In the United States, the FBI's National Crime Information Center logs roughly 600,000 missing-person reports each year — about one for every 550 people in the country, every year. Most of those people are home within hours; the great majority within a day. But not all of them. As of March 2025, the U.S. NamUs database held more than 26,000 open missing-person cases that had never been resolved, and roughly 15,000 unidentified-person cases in parallel.

These numbers describe a paradox at the heart of disappearance: the difference between found and not found almost always comes down to how quickly someone knew where to start looking.

The geometry of being lost

If you walk in a random direction at 3 km/h for one hour, you might be anywhere inside a circle 6 km across. After two hours: 12 km across, an area four times as large. After four hours, sixteen times as large.

This is the geometry that organizes every search and rescue operation in the world. Search area grows with the square of time. Every hour that passes between the last known location and the start of the search makes the job four times harder than the hour before.

Robert Koester, who built the International Search and Rescue Incident Database (ISRID), has spent thirty years cataloguing where lost people actually end up. ISRID now contains over 200,000 incidents. The headline finding from all of it is dull but absolute: more than 90% of lost adult hikers are found alive if found within 24 hours. The curve drops sharply after that.

What ISRID also makes clear is that "lost" doesn't behave the way TV shows it. Lost adults rarely walk in straight lines; they tend to circle. They tend to travel downhill, not uphill. They tend to follow drainages and trails. The category of lost person — hiker, child, dementia patient, despondent adult — predicts the search pattern more reliably than the terrain does.

Modern SAR teams are guided by exactly this kind of probabilistic model. They don't search outward from the last known point in concentric rings; they search the regions where Koester's data says this category of lost person is most likely to be. The number of hours since the last fix sets the radius. The category sets the shape.

Children with autism

A 2012 study in Pediatrics found that 49% of parents of children with autism reported at least one elopement event — a child wandering or running away from a safe environment without warning — between the ages of 4 and 10. Of those who eloped, a third had been gone long enough that police were called.

Drowning is the leading cause of death for children with autism who go missing. Bodies of water are attractive terrain for autistic wanderers; they're also where SAR teams now search first, by protocol, because the data is unambiguous.

The single most useful piece of information in those searches is the last known position. Not a 24-hour-old location. Not "she was at home." A timestamp and a coordinate from within the last hour can compress the search area by more than two orders of magnitude.

Dementia

The Alzheimer's Association estimates that 60% of people living with dementia will wander at least once. For many, it becomes a pattern. The medical literature is alarmingly consistent: if a wanderer with dementia is not found within 24 hours, the risk of serious injury or death rises to roughly 50%.

The reasons are physical, not behavioural. Dementia patients don't tend to seek shelter. They don't tend to ask for help. They tend to walk in a constant direction until something stops them — a fence, a body of water, exhaustion. SAR teams trained in lost-person behaviour now treat a missing senior with dementia as a higher-urgency case than a missing healthy adult of the same age, almost by definition.

Here too, the most valuable single artifact is a recent location ping. SAR coordinators have told local press, over and over, that the difference between a same-day find and a multi-day search is almost always a smartphone whose last-known location was less than an hour stale.

Wilderness: the golden day

The "golden hour" is borrowed from emergency medicine; in search and rescue it stretches to a golden day. ISRID-derived curves show that for healthy adult hikers found within the first 24 hours, survival rates are above 90%. Between 24 and 48 hours, survival probability begins to fall steeply, driven mostly by exposure — hypothermia in cold rain, dehydration in heat, falls in rough terrain.

Weather is the multiplier. A lost hiker on a 20°C day with no rain has time. A lost hiker in a winter storm has hours, not days. SAR teams allocate resources accordingly: more searchers, faster mobilization, narrower search areas when conditions are hostile.

The privacy tradeoff people actually make

There is a real reason consumer location-tracking apps have a bad reputation. Life360, the largest of them, was caught in 2021 selling precise location data from tens of millions of family members to data brokers. Apple's Find My, opt-in by default, is so reliable that abusive partners have used it to stalk spouses. The fear that a sharing app will turn into a surveillance instrument is not abstract; the history is full of examples.

But the alternative isn't private — it's opaque. A teenager who refuses to share location with a parent isn't safer; they're harder to find on the day something goes wrong. A spouse who can't tell whether their partner left the airport isn't liberated; they're guessing.

The actual question is not whether to share, but with whom, how precisely, and on what terms. A precise live location to one person; a city-level "outside" to another; nothing at all to most. A protocol where the publisher of the app cannot read any of it, even if subpoenaed. A history that lives on the recipient's device, not in a corporate database.

What FreeBusy does about it

FreeBusy is built on the premise that the cost of sharing is too high because the sharing is too coarse and too public, and that the data above — every minute matters, search radius grows with the square of time, recent location is worth more than detailed history — is best served by an app that takes consent and granularity seriously.

We can't help find someone the first time something goes wrong. What we can do is make it possible for the people who love you to start the search with a four-minute-old position instead of a four-hour-old guess. That is, on the curves search and rescue teams actually use, the difference between found and not.

See FreeBusy →

Notes & sources