Loading ...
Sorry, an error occurred while loading the content.

RE: [aprsisce] feature request mid point

Expand Messages
  • KK4DFZ
    At sea, the problem is known as the Closest Point of Approach (CPA). Important in judging the risk of collisions, but it is a much simpler proposition if you
    Message 1 of 5 , Jun 28, 2012
    • 0 Attachment
      At sea, the problem is known as the Closest Point of Approach (CPA).
      Important in judging the risk of collisions, but it is a much simpler
      proposition if you don't have to stay on a road.

      There is some nautical freeware called SeaClear II available at sping.com
      for anyone that would like to see what one implementation looks like.
      You'll have to provide an NMEA location feed to simulate your own movement
      and find an online AIS feed to simulate other traffic. I think there is one
      available for San Francisco Bay but don't have the link handy.

      I run both SeaClear and APRSIS32 on my car laptop sometimes as we have OSM
      and nautical chart coverage of the roads and bridges I drive. Kinda funny
      to be warned that your Explorer is in danger of collision with an oil tanker
      that's 100 feet below the roadbed.

      I'd be surprised if anyone had implemented a similar solution for the road
      atlas based topology used in automotive GPS, but for rural and freeway
      driving, the simple marine CPA may have some utility and not be too hard to

      -----Original Message-----
      From: aprsisce@yahoogroups.com [mailto:aprsisce@yahoogroups.com] On Behalf
      Of James Ewen
      Sent: Thursday, June 28, 2012 8:48 PM
      To: aprsisce@yahoogroups.com
      Subject: Re: [aprsisce] feature request mid point

      On Thu, Jun 28, 2012 at 5:58 PM, Steve Daniels
      <steve@...> wrote:

      > It's a help to make a decision, not a be all and end all to things.
      > I don't blindly follow directions and neither do you I expect James.
      > You get help from the tech.

      Yup, but I was just playing the opposite side of the fence... Driving on a
      straight road towards each other makes the job of figuring out a mid-point
      fairly easy.

      Without any routing information, all you can do is draw a straight line
      between the two points. You could take the apparent closing speed of each
      unit and try to come up with a mid-point based on those numbers, but how
      good would that estimate be, and would it be any better than what you could
      do just looking at the map?

      Around here it would probably work out most of the time for two people
      driving towards each other because our roads are fairly straight.

      Take this example though... I'm driving northbound on road A, and you are
      driving westbound on road B. These roads are perfectly straight, and
      intersect perpendicular to each other at point C. We are both travelling 60
      km/h, and are 30 km from point C. Simple observation would suggest that the
      most useful meeting point would be at point C, and we would meet there in
      1/2 an hour.

      If you calculated a straight line between us, the meeting point would be
      halfway along the hypotenuse 21 km from each of us. The velocity made good
      towards each other along that hypotenuse would be 42 km/h, even though we
      aren't moving in that direction, and that VMG would decrease as the angle
      towards the location increased away from our direction of travel,
      effectively making longer and longer to get to the meeting point until we
      end up moving away from the mid point. Of course with the calculations
      happening as we drive towards each other, the mid point will move at a 45
      degree angle towards point C, at which time we will have arrived at the
      meeting place that we already knew would be the best location for both of

      As the angle of intersection between the two roads decreases, the apparent
      midpoint would get closer and closer even though point C would still be the
      best probable meeting point without taking either of us out of our way.

      Hopefully this example shows why simple mathematical calculations might not
      be all that helpful with no knowledge of the underlying topology of the

      It's a similar problem when I search for a POI on the GPS. I tell the GPS to
      look for a gas station for me. It tells me the closest one is 2 km away.
      Great, I tell it to take me there, and the GPS starts telling me to travel
      away from my desired destination because the gas station is in the opposite
      direction. However, the next closest gas station that was 3 km away is right
      on my desired route, and is much more convenient for me, but since the GPS
      has no idea where I intend to go (I'm sitting in a parking lot playing with
      the GPS with no route selected), all it can do is suggest gas stations based
      on proximity.

      I'm not saying the concept is bad, but the amount of information needed to
      make an informed decision based on all available criteria is very complex.
      It might not seem like it, but when you break it down into steps, it can be
      very complex. That grey goo between our ears does some pretty fancy
      calculations for us some days.



      Yahoo! Groups Links

      No virus found in this message.
      Checked by AVG - www.avg.com
      Version: 2012.0.2180 / Virus Database: 2437/5099 - Release Date: 06/28/12
    Your message has been successfully submitted and would be delivered to recipients shortly.