ModelTracker trundles along

[HankKauffmann]

I was out at the field today and snapped off a couple quick flights.

ModelTracker under formal race mode with pylon 1 start-light, and turns & cuts.
Flying starts at 2:10. ( The video looks better in YouTube full screen, Click the YouTube icon at the lower-right. )


ModelTracker under practice mode with pylon 1 and turns & cuts.
Flying starts at 1:35


Thanks to Tom Dooly of Phoenix who dutifully mines the large data downloads after each practice session looking for trends. His analysis and insight in identifying where the effort needed to be spent has been valuable. The system has made steady progress under his watchful eye. I am looking forward to testing in multiple planes next weekend.

[RSmith]

Looks good Hank and Tom. I notice that the light comes on when you break the plane of pylon #1 and stays on for the time you are past the pylon. This gives the caller a good indication of how far past the pylon the plane flew. Is lap counting now functional too? Laps and P1 could reduce six to eight people required for the course.


Randy

[HankKauffmann]

Yes, there is no need to enforce the 2 second turn-light rule. That rule on Judgeman was to eliminate variances in button pushes by the workers at P1.With the human element removed, the system can help the racers by giving them useful feedback. For some that will be like giving them enough rope to hang themselves.

I just checked the data from the plane's tag and internally it was doing everything just fine. I'm changing the lap-count display to use the same data source as pylon 1. It's less prone to problems.

Next test is to simulate 4 plane races. For that test I'm have extra plane tags on the ground as well as a plane in the air. This will stress test the system's latest radio code. I've flown with two tags on a plane in the past, but with dual antennas on all tags I'm running out of places to put more antennas.

Hank

[KRProton]

Wow!

This would be a monumental gift to the pylon community!

Either way, thank you Hank (and everyone else who has worked on this too such as Tom Dooly) for all of your efforts and untold hours.

Tim Lampe

[rs-ssfan]

Hank, things are looking great on the model tracker and look forward to the multi-plane test results.

I'm curious about the costs associated with the new system. Will it integrate with current judgeman systems or will clubs have to buy a whole new set up? Could you give us a ballpark cost on upgrading/purchasing new? At one time I'd heard the transponders were going to be in the $100-ish range, is this still the projected cost point? I'm assuming each plane will need their own transponders. Will they be internally or externally mounted?

I know this is putting the cart before the horse but for a small club like mine, I've got to do some planning a head for any major purchases. So I would appreciate any insight you could give me on this subject.

Thanks,
Lonnie

[tomdooly]

Some lap plots from the 1st video flight...

The plots show the path of the plane, the location of the six course transponders (Master, Aux1, Aux2,
Pylon1, Pylon 2, and Pylon 3), and an approx. position of the camera shooting the video.

Lap 5 shows the cut at Pylon 1.
Lap 6 shows the next lap with NO cut.
Landing shows the plane landing after the race.

Onward and Upward!
Tom

[HankKauffmann]

Lonnie, The short answer is the plan includes a bridge between Judgeman and ModelTracker systems. Hardware-wise the bridge has transponders for both systems. It can be display-less or include any of the ModelTracker displays such as the number display with punctuation or the "Race Summary Display" you've see at the NATs.

The system will cost somewhere between zero dollars and the amount I personally sunk into it The cost of each tag will be reasonable, however the total cost once we determine MOQ (minimum order quantity), not so much.

We in Calgary had 100 tags made up and even with a matching government grant our club coffers took a serious hit.

[rs-ssfan]

Thanks Hank. Look forward to the next set of results.

[RCjunkie]

Next time please show what the screen look like in a race

[HankKauffmann]

I could do that, but you won't like it , it's just a diagnostic program I use to operate the hardware with.

There's the old line about the dancing bear, "It's not about how well the bear dances, but that it can dance at all". I've been focusing on the hardware aspects of the system and not so much the laptop side of race scoring.

The hardware does do scoring as the planes log everything and shares with the rest of the system all their race events along with associated times for laps, turns, cuts and passes.The master captures these events and forwards them to the computer for posterity. The master also uses that same data to determine the finish order and times. These are displayed on the main number display while less interesting than watching turns at one but I can video that display when we have a two man race.

Your implicit point on a full score keeping system on the laptop is taken though. I haven't yet thought about it, but my initial idea is to adjust the JudgeTimer program to accept and display ModelTracker race data.

Kevin Umbach of Edmonton made a race scoring program that reads the JudgeTimer data file to produce a WiFi / web based leader-board. I used it when we were in Dubai and it was great. It does need a computer literate guy to ride shotgun on it and make adjustments at the end of each race to correct finish order and fix times. His program has some wonderful features including the ability to import race matrices and pilot lists.

With the automation of ModelTracker the amount of mollycoddling and editing these scoring system require should be reduced. Programs like Kevin's should be easier to implement.

The plane tags along with having unique IDs can be programmed with pilot names, numbers, lanes, etc. This should make it easier for scoring systems to properly attribute points or match up heats with the planes that show up at the line. It will also eliminate the need for stickers and preassigned lanes. Nothing is stopping you from having an event where everyone races everyone else.

Finally the current system is focused on pylon 1 turn and lap-counting as these are the most difficult, problematic, and man-intensive tasks. With ModelTracker we can eliminate the timer judges and turn-button pushers at the NATs and other major events. We'd just have judges to issue cuts at pylon 1. At local races you'd only need 3 or 4 guys to man the course.

[fizzwater2]

What would the chances be of moving away from 2.4G wifi for the wireless links between computers, to something perhaps a little more tolerant of antenna position, etc?

At some of the fields in this area where we race, the judgeman system has issues with reliable connections. We've had to rig up extension cables, poles for antenna mounting, etc., and even with that, the wifi links aren't always 100% reliable.

[HankKauffmann]

There is a newer and longer range 2.4GHz tech from the same company that I suspect is a drop-in replacement but you'd/we'd have to replace all the modules in all stations. I think they also have a 900Mhz module, you'd have to change antennas too.

A recent revelation after this NATs was that the thicknesses of the enclosure relating to the installation of antennas had caused the antenna to not screw in far enough to properly engage contact with the center pin. This could explain lots of weirdness we've seen over the years.

[ceandra]

Yes, that was a key discovery by Tom. He worked his butt off trying to keep up with system issues. This connector interface was subtle, but critical. When he filed down the connectors, our problems pretty much were gone.

Chuck
11H

[iamtom]

The Tom is Tom Melsheimer. I am very hopeful the is the answer to many of our communication issues. I lan Too
To modify our system before the Muncie race later in August to see if we are on the right track.
Forever hopeful.
Tom Scott

[RCjunkie]

What I was looking for what was happens to the screen as you put laps on it