Rerouting

Ask for help, find useful guides, or help others out with their issues.
Post Reply
Brian
Posts: 51
Joined: Thu Mar 08, 2018 1:45 pm
Location: Altamont, New York, USA

Rerouting

Post by Brian »

I am a very early K1 owner. While HH struggled early with their software, I've been extremely impressed with how far they have come. So impressed that I immediately ordered a K2 when given the opportunity.

Today I was out for a ride. The weather was supposed to get better during the ride, but never did. I got to a point where I had 16 miles to go on my planned route or 10 miles to go if I shorted. Since I couldn't feel my fingers or toes, I chose to short. I knew the way home without nav, but I figured that with this fancy nav computer, I would simply reroute to my house. I've never had to reroute before, so I wasn't sure how to do it. I pressed on the screen (remembering this from some post) and a purple dot with four cardinal direction arrows appeared. I moved the purple dot over my house and said "go" - or at least I thought I did. Anyway, the K1 was definitely taking me to my house but basically backtracking the way I came - not the quickest way home.

So I'm asking this esteemed forum what the secret is to doing a quick/easy reroute to your starting point. A "return to home" option would be awesome, but life is never that easy - usually...

Thanks.
User avatar
Psyclist
Posts: 66
Joined: Fri Jul 06, 2018 1:26 am
Location: Vientiane, Laos

Re: Rerouting

Post by Psyclist »

As far as I've seen, that's actually the way it usually works when rerouting - it finds a pretty good approximation of the quickest way. Maybe it got confused on your map, idk.
It looks like the Karoo's internal routing algorithm is mainly optimized to be quick to find a solution. If you compare it to say OsmAnd, the latter usually finds a good solution but it also spends orders of magnitude more time to find it. There haven't been any revolutionary new algorithms for the routing problem in more than half a century (the state of the art is actually still some variation of Dijkstra's algorithm from 1956!); every real-world solution is simply a compromise between finding the optimal route in impractical time or a crap route very quickly. It's a safe bet that HH's compromise is simply more on the side of calculation speed.
Post Reply