Hi!

I decided to do a "Test Drive" of some well known and readily available Software Routing Programs. This is MY perspective of Sailplanner, Expedition, and Bluewater Racing, and criticism of any kind to my comments is more than welcome. I'm no authority on routing software, just a regular guy tring to sort things out. The purpose of this test was to see how they stack up against each other and point out some of their pros and cons. The course utilized for the test is the upcoming Ouzo Rally, the polar file (First 407 basic txt file) and the grib file (weather_36_global_gfs_20100715_1138.grb) used are the ones from brainaid's site. So all programs are "fed" with the same data. As far as I know Sailplanner's weather provider is the same as SOL, with the caveat that SP's weather is interpolated in a different manner from SOL's weather, so we are not using the same grib file for SP's routing but it should be close. All polar files are EXACTLY the same. All races were started at the same time 7/15/2010 12 UTC. So with all the aforementioned caveats, lets get to the nitty gritty.

Sailplanner (www.sailplanner.net) in my opinion is the easiest one to use, but that user friendliness comes at the price of flexibility and a few advanced features. First in SP you don't have the freedom to choose your grib files, they are dowloaded directly from sailport's server for the area of the world where your course is. Setting up waypoints in SP is pretty straightforward as it will parse almost any format of degrees minutes and seconds which is pretty cool, and its graphical user interface beats the others in my opinion. As usual on the first try SP will route over land, then it will remove any gridpoints that are over land and proceed to re-route. See attached pictures. For the Ouzo Rally course SP gave a ETA of 7/16/2010 02:50 UTC roughly 14.8 hours. One thing SP sorely needs is to have the ability to output the routing results into a text file, and the ability to save prior routes. When comparing the attached pictures you will see that SP is up to par with the other two programs, specially when routing around land.

I got a 30 day trial license for Expedition (www.iexpedition.org), the program costs a whooping $1,250.00 US. During the trial period you can test all the features of the program, it is NOT a watered down version. Most of these features apply to short course stuff which we don't use on SOL. Setting up waypoints is not as intuitive as SP but its not that bad once you get used to it. Since Sailplanner's Ultra High resolution grid longest dimension was roughly 1.2 nm I set Expedition to a routing resolution of 1.0 nm. On its first try Expedition quickly routed above land, (as expected), even when selecting the avoid land option!! For the second try I added some exclusion zones (i.e. land) and this time Expedition gave an acceptable route with an ETA of 7/16/2010 02:42 UTC (14 hours 42 minutes 38 seconds).

Bluewater Racing is freely available from www.bluewaterracing.com . It's not as user friendly as Sailplanner, so I suggest you thoroughly read the manual (twice) before starting to fiddle around with it. Once the waypoints were setup I proceeded to optimize the route with a time resolution of 0.1667 hours (10 minutes which is the same as SP) a grid size of 1 nm, a distance resolution of 60.0 nm (default) which refers to some flat earth approximation mumbo jumbo... and a Max Range of 10 nm (tells the program the allowance for the longest section of routing). As expected it routed over land. Getting Bluewater to not route over land is harder than in Expedition or Sailplanner. A few marks (one at Vouliagmeni, one at Charakas) had to be put into place to force Bluewater into an acceptable route. The output on the 2nd try (blue route) was suspect, since the prior route (red) favored a more northerly approach to Kythnos. I added a couple more marks, one at southern Kea and one at northwestern Kythnos. This 3rd optimization gave a slightly faster route (purple route) than the blue route with a northerly approach to Kythnos. The Blue Route had an ETA of 7/16/2010 03:06 UTC, the Purple Route had an ETA of 7/16/2010 02:53 UTC for the sake of comparison, the Red Route (useless since it goes over land) had an ETA of 7/16/2010 02:49 UTC.

Other observations

It seems to me that Expeditions weather interpolation scheme is the one that most closely resembles SOL's weather interpolation. All of these programs are intended as IRL tools so why doesn't Sailport use this same interpolation scheme on Sailplanner is beyond me. Sailplanner and Expedition automatically compute the best VMG angles for upwind and downwind from polar data. Bluewater doesn't. In Bluewater you have to TELL the program which are the best VMG angles for pure upwind and pure downwind for each windspeed. The second entry on each line of polar data tells Bluewater which is the best upwind VMG angle. The second to last entry on each line of polar data assigns the best downwind VMG angle. From the polar file we can see that Bluewater thinks that 52 degrees is the best upwind VMG angle for all windspeeds and 150 degrees is the best downwind VMG angle. Experience tells us that is is totally wrong. Is this the reason for the highly suspect blue route? I think so, but more on that later.

Conclusion

All of these programs will give pretty much the same output given the same data. The uncertainty on the next weather forecast is greater than the differences in output from one program to the next. So anyone using these tools will get decent results as long as they don't blindly accept the output. I usually treat the output as if it were wrong, I analyze it and if the analysis makes sense, then I trust it.

Hi Exmeromotu,

I use prettier colours for my routing
Attached is an example.......

No you dont sol........NZL_undercover does

Here is a race day followup to the Software Test:

Sailplanner has an ETA of 7/17/2010 22:07 UTC

Bluewater;

All the routes pictured in Bluewater are within 1 minute of each other. I played around with the grid dimensions to see if I could get it to route without going over land. I finally did (Red Route) with a grid dimension of 0.25 nm! No additional marks to "force" the route were necessary. I also fine tuned the polar data so that the program knows which are the optimum VMG angles. ETA's range between 7/17/2010 22:01 UTC and 22:02 UTC

Expedition:

I drew some exclusion areas (land) utilizing the coast profiles from the SOL chart. ETA is 7/17/2010 21:59 UTC using the middle of the line as finish point. The three legs from Kythnos to Paros are for the different arrival points.

Conclusion:

Once I fine tuned Bluewater's polar file and grid dimensions I could get it to route avoiding land. So all the software programs could handle the land obstacles decently, albeit with some user help. Sailplanner needed the least user input as far as land avoidance is concerned. Expedition needed the most user input to avoid land, and Bluewater was somewhere in between since you had to fiddle around with grid dimensions. It should be noted that all these programs use different Charts with different coastline resolutions, so the routing over land problem would take a whole new perspective if we were doing this race IRL. Once again all software programs give more than decent results their differences resulting in mainly the way they interpolate the weather. Some human intervention is needed in all of them to fine tune results.

On a "Price/Performance Ratio" well Bluewater is free , Sailplanner costs $89.00 and Expedition costs $1,250.00.

On a "User friendliness/Performance Ratio" I would rate Sailplanner first, Expedition second and Bluewater third. Expedition and Bluewater have more flexibility and features than Sailplanner though. Attached are Race day pics enjoy.

In software that supports it, a "trick" that's a good look into what is going on for those wanting to learn from the software: Instead of routing to both ends/middle of the finish, turn on forward-isochrones for a visual representation of the fastest target.

This works even when the line is long and an intermediate point may be fastest.

Cool Trick! Didn't think of it. I was doing it "the hard way", nevertheless, I was aiming for the "fastest point" of the line, and the Isochrone trick confirmed it!! Nice!!

For those interested...

I raced the Ouzo race completely on DC's (no human interviention) using Expedition. I didn't hit land and I arrived 3 seconds earlier than scheduled... however 39 boats sailed faster than I did. The top boats beat me by 30 seconds!!

Granted, it was a short race and not very router friendly (lots of obstacles) but nevertheless, a pad, pencil & longhand calcs with a little bit of knowhow will produce awesome results too!!

Thank you Exmer for sharing with us this interesting comparison

(Unquote)


I'll do you one better: racing the Scurry timed race on 1 manually entered command, one to get the boat going. All other commands are input by a tool. Can you guess?

Philip

Looks pretty slick Schakel! Cool!