On Sat, 26 May 2012 14:30:58 -0400, Bob Flumere wrote:
> My Garmin 2450 gives a route from point "A" (Home) to point "B".
> Fine, and then I would like to go from point "B" back to point "A"
> (Home) on the same roads, but, my 2450 gives a somewhat different
> route almost every time.
> Settings are "Shortest Time".
> Why is the reciprocal route sometimes (almost always) different in
> some way from the original route?
> Anybody else experience this phenom?
"Shortest time" means shortest time. Not shortest drive. The 2450 is a
trafficTrends unit, which means it takes notes about routes it sends you
down. Which means it learns which intersections take a long time to get
through at that time of day, which left turns are no big deal and which
take a long time get through, which roads that are posted as 45 MPH
really let you through at 45 MPH and which are congested enough that the
average is 30 (again, at that time of day), etc. It'll also try
different routes to see if they're better or worse than the detailed
information it has about the "obvious" fastest route. And it's picture
of where you drive and how much you speed routinely will also change
over time so that it can get better at selecting "fastest" for your
particular driving style and patterns.
As an example, my unit knows that the "obvious" direct route to the
interstate isn't the best. It's a six-lane US highway posted for 45
MPH that happens to also have six traffic lights on the length of it,
and three stop signs to get to it by the obvious route. If it takes a
zig-zag route out from my house that's "out of the way", it can skip
two stop signs and the one that it does encounter is a right turn. In
the afternoon, that 6-lane road's traffic lights are synced to traffic
flowing *into* town, not out to the highway. So in the afternoons, it
sends me down a 30 MPH semi-residential side street (instead of the 45
MPH main drag) that has comes out between the second-to-last traffic
light and the last one before the highway. It took me slavishly
following its directions, regardless of how dumb they seemed, for about
three weeks before it started working out to be faster ALL the time
instead of just most of the time, but now, yes, it really DOES know what
its doing better than I do most of the time, for that path.
"... I've seen Sun monitors on fire off the side of the multimedia lab.
I've seen NTU lights glitter in the dark near the Mail Gate.
All these things will be lost in time, like the root partition last
week. Time to die...". -- Peter Gutmann in the scary.devil.monastery