nah, didn't really expect it...

[cellio]

It was unlikely, but it would be cool if MapQuest (or some similar service) provided a path-finding option that minimizes the expenditure of kinetic energy. For, y'know, pedestrians who want to take hills into account. The only customization option I see is "avoid highways", which I was going to do anyway. :-)

Oh well. I will just have to rely on my innate knowledge of the topography involved.

Comments

[eub.livejournal.com]

I was thinking about how to do this, once. Minimize some weightable sum of horizontal distance and total vertical distance gained? But then I started trying the corner cases: scale a five-meter cliff, or walk fifteen meters up a thirty-degree grade? Hmm. Maybe some steepness term in there too. I like the "energy expended" idea, if it has a good model somehow.

Scanned topo maps are available online, for lots of the continental U.S. anyway. Seems somebody could parse them into nice heightfield information.

[cellio]

I initially thought about it in terms of veritcal distance but ran into the steepness problem, which is what led me to cast it in terms of energy. I don't know how easy that is to model, though. But maybe it could be parameterized -- minimize total vertical climb while specifying a threshold for steepness. Someone in a wheelchair, for instance, is going to have different constraints than a pedestrian does anyway, and a bicyclist can probably accommodate greater steepness without trouble. (It's all about channeling the energy from that running start. :-) )