Hello! We are students working on a project to reduce the number of missing people in national and state parks. Based on the feedback from our recent post, we have developed an updated prototype and would greatly appreciate your thoughts on it.
At first, we treated this as mostly a tooling problem (bad maps, no cell service, too many apps). But after thinking more about it, a big part of the issue is actually psychological. This connects to the preparedness paradox. The people most likely to get lost are often the least likely to prepare ahead of time or actually use a safety app when something goes wrong. Overconfidence and “I’ll be fine” thinking often win over preparation.
That made us question whether a standalone safety app or a pure education system really works. Education helps, but people do not want extra work before a hike, and most learning tools are easy to ignore. At the same time, tech alone does not really fix behavior either.
Current direction (Atlas, reframed):
Instead of “an app you have to learn,” we are trying to think of Atlas as a tool you just naturally use.
At a high level, the idea is to take the most useful parts of existing outdoor apps and combine them into one free, offline-first system.
So instead of jumping between different apps, you would have:
- Trails and maps (AllTrails-style)
- Location and tracking (Garmin-style)
- Basic survival info (first aid, knots, decision making)
- Park specific warnings and info
Right now, all of these are split across different tools, which makes it harder for people to actually use them when it matters. The idea is that by embedding everything into the trail experience itself, the right info shows up at the right time instead of relying on people to prepare in advance.
What we want feedback on:
- Can this actually help with the preparedness paradox, or does it still fail the same way?
- Is focusing just on trails more realistic than trying to cover everything?
- Are we still overestimating how much people will use something like this in real situations?
- How well do you think this addresses the problem compared to our past 2 prototypes?
