Printable View
if you track the FCC ID you can probably find the frequencies. i'm betting it works a bit like FRS.
The connectors are a standard Deutsch design for wet applications.
the other components on the display board are switches that go under the white buttons on the front of the unit
I did some research last night and found the info on the chip they are using. It's a packet chip transceiver that's pretty sophisticated and CHEAP and it operates in the 'free spectrum' frs area around 900 MHz. It still raises a question in my mind as to the latency of comms to/from the car. It's still going to take, either a distributed antenna system at the track, or a VERY HIGH transceive antenna (like HELICOPTER high) to reach dead spots on a lot of tracks. The data rate possible is also pretty low. I'd be really interested to know more details, but couldn't find ANY 'specs' on the Flagtronics website.. most info is 'call for more info'.
Makes me wonder if anyone has actually TESTED what can be expected from this system.. or just listened to the ads. I haven't seen any ad info that addresses the time frame of the comms.. or the hardware required to 'manage' the cars at Race Control. Even though race control is usually 'pretty high', it would not be nearly high ENOUGH (IMHO) to reach into places like the gap between flag stations(FS) 4 and 5 at Road Atlanta, or FS 9 - 11 at Mid Ohio.. or FS 8 - 12 at Road America. Although the cars should EVENTUALLY get flag info, it could take quite a few seconds for that flag display to show up in cars in those areas .. unless they have figured out a way around that issue. SO.. a RED FLAG at Elkhart issued when a car is in the carousel might not get displayed until the car comes out of the 'back side' up to around FS 13. Or a car at Atlanta going into T3 might not get 'the word' until he got to the top of the hill at T5, so if the incident is between those places (right at where the car is headed) he might not get notified before it's too late.
That also raises the issue about DETERMINING PUY via this system and KNOWING exactly WHEN 'the car' actually had a yellow displayed IN HIS CAR.
I'd like to hear some data about real world testing where the real world KNOWS exactly when a flag command is issued at Race Control vs exactly WHEN that command is received and acted on at various places around the track.
I'm betting there could be quite a few seconds before the car gets the info in some cases ..
A car traveling at 100 MPH is covering ~145 feet (or ~12 car lengths) per second. I doubt the data rate of the packet system can keep up with the GPS position of even ONE car at that speed.. much less 20 of them. SO.. 'verification' of lap time or PUY seems pretty suspect to me.
Of course.. I COULD BE WRONG, but when was the last time THAT happened ?? :D:D
Steve
I'll make some more guesses here. The reason I thought it was a FRS type system is that it needs comms to be channelized per car even thought there's not going to be enough bandwidth by conventional standards and the system has to be easily set up by race officials. So either you'll have a number that has to be presented to the officials like a transponder number and entered into the system before a session, or each in car system will have two "channels" - one code for an open broadcast and another coded for the individual transmitter. That way the track system can "ping" each car and get a response as to who is out there.
900 MHz can go a long way. It sits at the edge of a band with less atmospheric absorption and somewhat less natural noise levels. The track system can be high powered transmit, receive is somewhat less important. Each transmission is a small burst with high peak power/short duration. Using some very conservative and generalized estimates, if you consider the information bandwidth to be on the order of 100MHz, and you're sending sync and a hundred bits of info plus error correction, that gets done in 3uS. For green/black all/red/FCY, it's a single transmission to all parties. Boom.
Acknowledgement is a different matter, but they have repeaters that can be placed along the track and relay. They need LOS to each other and the main base, so you just put them on a pole somewhere. Even multiple repeater delays are only going to be in the range of tens of microseconds. Even if I'm off by two orders of magnitude, its still 50 times faster than you are in recognizing it.
Typical GPS update rates even for aircraft are on the order of a 10hz. You can get somewhat higher at great computational and power expense, so its simpler to couple with an IMU and extrapolate if you need more position accuracy above 10hz.
I think the more worrisome aspect is "cold start" on the vehicle units. Without a cellular/internet connection it takes a long time to download ephemeris data in order for the system to find the satellites and produce a solution, although once that's done it's usually good for hours, if not days. So you won't want to show up on grid without having "warmed" up the GPS first, unless it gets that data from the flagtronics ground station in a general broadcast message every so often.
I was surprised that the GPS aspect is an option, because if you don't have that all you get is green/black all, red, and FCY. Considering how rare those events are, its a lot of effort for not much gain, so I'd expect the requirement to be for the entire system.
It appears the main reason the display is so large is that they are displaying a character to indicate the flag color by not illuminating LEDs, R,G,Y etc. Probably to assist with a certain amount of color blindness. If they are displaying passing and course condition flags, a strip display would require alternating segments of red/yellow and blue/red, as well as something for the meatball. How are they implementing black? Some interesting human factors here (and I doubt if anyone is doing real human factors work, most likely just a collection of opinions like the basis behind the FIA rain light).
Tried to download the user manual but the link is broken.
Steve,
I don't have the answers to your questions, just another question: How are local yellows triggered?
It sure sounds like everything goes through Race Control. There's nothing to keep the local workers from throwing a flag while that process works through, but giving them access to the system probably complicates things A LOT.Quote:
Originally Posted by Peter Olivola
I was just thinking that a different way to implement this would be a transmitter at each corner station with a coded signal for each one, and much lower power. The marshals would be in control of it, much like flags now, and due to the short range, it would only trigger when you came within sight of the station, and drop when you got the signal from the next station not displaying that color.
This would likely INCREASE the need for corner workers though.
There are a bunch of YouTube videos now. In one of them, the annotation makes notice of a local yellow thrown, and it's a few seconds before Flagtronics displays it - it was thrown when the car was just entering the braking zone but he's almost to the apex before it shows up on his display. So you can safely assume this is the communications delay between the workers reporting yellow, race control acknowledging it, and someone there operating the software.Quote:
Originally Posted by TimH
I don't think it would be all that difficult to give corner workers control of it - they already have a handheld tester that could probably be re-purposed into that role - but it would require another person to operate. I don't think you could watch, flag, and tap buttons at the same time.
And apparently you will need to report your "transponder" number prior to the event, vertical orientation of the antenna is critical, and they implement black via alternating panes of white with black text and off with white test. Checkers appear to be purple.
LOCAL yellows are triggered with FLAGS from the corner station by HUMAN communications visually... HOW does that get into the FT system?? Good question. I THINK FT is really ONLY for FULL COURSE situations, but I don't really have ANY IDEA how they might intend to implement it. THAT's the reason I asked the question about REAL WORLD TESTING. I KNOW it's 'been used' in a few events.. but the only comment I can recall is from Bob Wright and .. if you/he is IN THE CAR and 'sees a flag' on FT.. there's no way of knowing WHEN you/he saw it relative to WHEN it was initiated in the first place.Quote:
Originally Posted by Peter Olivola
I see that as a BIG HOLE in the entire MAGIC PROCESS that is going to solve ALL of our FLAG PROBLEMS.
OTOH, I can see that as CAUSING a WHOLE LOT MORE issues where the system 'initiates' a flag condition, but the car does NOT receive, recognize and DISPLAY that condition until much later.. yet RC would EXPECT the driver to have responded 'immediately'... when he was really not notified by FT until 15 seconds later.
Rick's comments about 'local interfaces' are well received, but THAT is the question I was asking about 'distributed' antennas for comms to the cars. When a corner worker SEES A CONDITION that calls for a flag... and then SHOWS that flag.. HOW does that 'trigger' the FT flag condition in the car?.. especially with FEWER F/C personnel which is at least PART of this problem .. not withstanding the 'overly aggressive idiocy' routinely exhibited by us 'inattentive drivers'.
And .. YES.. if you read the Runoffs Supps instructions for installing the system, they REQUIRE all drivers to NOTIFY .. someone (at tech?).. and supply them with THEIR specific numeric ID for the FT system.
Steve
So we can assume there is some latency with the system when an incident first starts, but when do the majority PUY really take place?
This system is not promoted as an instant responder. Some eyes will always need to be on the track: corner worker, race control tower, maybe video, or the Flagtronic noting the abnormal GPS location/speed.
This is racing, **** is always possible when you are an early arriver to an incident.
Brian
Actually, this system *IS* being promoted as an INSTANT RESPONDER. If RC intends to use this for PUY penalty application THAT is a major requirement. We are truly talking MILLISECONDS here .. between the INSTANT when a worker pulls out a flag and the TIME drivers are EXPECTED to Recognize it and STOP doing what he/they was/were doing <1 second ago. If the system latency is ... even 1 second (or more) in ANY case, then it CANNOT be used to JUDGE whether or not a driver properly responded to a given flag condition.Quote:
Originally Posted by Hardingfv32
How many bits does it take for FT to send each car's GPS POSITION (considering that basic GPS info is still only ~10 Hz? .. as Rick indicates).. then @ 100 mph, we are traveling a little over a CARLENGTH per GPS update. We can have FIVE+ cars in a passing condition between 2 GPS updates and we are truly talking MILLISECONDS. It's true that the DATA CAN TRAVEL between 2 locations in microseconds.. but if you include packet 'confirmation acks' then, that starts to move into MILLISECONDS pretty quickly.. and then throw in some 'dead spots' and delayed resends to complete that packet info before the system will acknowledge and ACT on it... then we get to SECONDS + how ever many seconds of 'dead area' the car happens to be in at the time of initiation of the new flag condition. Consider the T2 yellow at Road America that cost a driver a National Championship just a couple years ago.
I don't see ANY WAY this system can be accurate enough to issue penalties from like it seems that SCCA is expecting to be able to do.. of, course that does NOT mean that they won't DO IT, anyway. :(
This is not like the Tennis 'hawkeye' systems that can call balls in or out automatically.. RC does not have 10 cameras on all cars at any instant. Tennis balls only get a 100 feet or so away from the cameras.. RC can be MILES away from the cars... and the GPS data that tells where the car WAS a few milliseconds ago .. as well as the 'previous locations' of the cars around it.. and add to that, that the specific location of the GPS positioning antenna is NOT SPECIFIED (just like transponder locations except for spec cars).. but the transponder data has become the 'defacto' indicator of just HOW CLOSE the finish was...
Well... I digress :-(.
The word has come down..and eventually we'll SEE how it all works out... how ever many of us are left by that time.
I'm getting too worked up about 'these new things' suddenly coming down from 'ON HIGH'.
I'd still like to see some REAL WORLD DATA about the latency.
Steve
Geeze guys, go back and read what I wrote and then watch a couple of the YouTube videos - the ones from FT on how its implemented and how each part of the system works and then watch an in-car to see the reality, and it will be apparent.
Unless this eventually morphs to be triggered directly by the corner workers you'll need to have one eye on the flag stations and one on the device.
I'm thinking this will be much more useful in closed cockpit cars where you have a lot of visual obstructions and for endurance racing. For amateur open wheel, the difference might be a couple of seconds in recognizing a FCY or all black.
it's 24 bits for latitude, 24 bits for longitude at 1m accuracy.
There are two levels of latency here, one for the system to transmit its position, say, 128 bits. It does not matter what happens at the race control receiver, in that the decision making process to understand the data is much longer than the reception of the data, display, etc.
The other level of latency has to do with the humans in the loop between the corner workers and whoever is fat-fingering in the flag at race control. Obviously, some people are going to be better at it than others.
Packet acq's? The contribution there depends entirely on the implementation, but in general we don't do acqs as much any more because its far faster to include forward error correction rather than to ask for it again.
As far as sorting what happens when multiple cars approach a corner station under yellow, I'd think this system is going to be no worse than the workers, especially if the corner is manned by only one or two.
I can see that in pro racing where lots of $$$ are on the line you don't want to put the series outcome in the hands of volunteer workers. They've largely fixed this on ovals, but road courses require a different solution.
I don't like the implementation of this at the amateur level for a few reasons:1) I'm not sure this will be a huge improvement over what we have, in amateur level open wheel cars, that some "culture stiffening" couldn't provide the same results for frequent violators. I don't think the SCCA has the backbone for that though.
2) it's just another thing I have to buy, install, verify, fix when it fails, etc, and its not engineered for the amateur open wheel environment.
3) when we finally don't have enough workers and this thing becomes not only mandatory but safety critical, then a failure of it will put you on the trailer for the rest of the weekend, so then I'll need an entire spare system (or there will no doubt be rentals).
4) when you have this thing it makes solving the corner worker problem less of an issue, and people will stop trying.
First, PUY reporting is just one one of the benefits of this system.
Again, when do the majority of PUY occur at this time currently? I would say the main focus is get speed control of the field at an incident. No one complains about stuff happening moments after an incident. That is just accepted.
There is no reason that a tolerance cannot be built in to the PUY reporting. If the tolerance is too tight, then you get unseasonable PUY reports. Also, if the system is actually making mistakes it is going to show up as too many drivers claiming innocence. Race control will need to learn how to best use the PUY reporting feature.
Brian
Brian, But if the Flagtronics is promoted as being able to reduce the worker count it seems kind of silly that it requires a worker to notify race control, race control to tell someone to press a button and then the magic Flagtronics gores into action.Quote:
Originally Posted by Hardingfv32
If it is not activated directly by the corner station for local yellows, it is counterproductive and will IMHO eventually result in a really incident due to the delay in notifying the drivers.
local yellows are triggered by the corner worker device.
https://flagtronics.com/collections/...40731848376480
saw that in a video but thought it was just a test device.Quote:
Originally Posted by BeerBudgetRacing
If that was what was used in the video there's waaaaay too much latency: https://www.youtube.com/watch?v=EV7azw3jFGA
This was the vie\deo I was commenting on, just noticed it was 4 years ago....most of the rest of the demo videos don't compare the appearance of an actual flag with the change in display.
Well, remember a couple of things.Quote:
Originally Posted by Rick Kirchner
- We can't see the point on track where the station is. Yellow begins at the station so that should be about 90 degs to the car. So even though we see the corner flag and they flash the message that they pressed the button, the car isn't under yellow yet.
- Theoretically the yellow should go off at the next (non-yellow) station.
- The corner workers device communicates with the master controller and that broadcasts to the cars based on their position.
At least I have to assume they are using the rules. ;)
It would be interesting to see if they start handing out penalties for PUY. I see it all the time.
Just thing of Star Mazda at ButtonWillow. How many times has a car gone off there and the station is waving yellow. How many people wait until the station at Sunset or Phil Hill to go back to racing? ZERO.
Good catch. I was assuming given the position of the car, that you would be yellow as soon as you see it, but that's not really the rule.... My brain was just looking at the physics and human factors of it without considering the details of the actual rule.
The specific software implementation would be interesting.
This sort of takes out the mentality of "racing to the yellow" because you won't see it first, it will suddenly just be there. There's a whole lot of mental gymnastics that could go on if both live flaggers and this device are operating simultaneously.
If I have missed this point regarding the device already being discussed, I will apologize in advance.
We all have raced in the rain, sometimes in heavy downpours. I recall a race in a monsoon at CMP where post race I weighed in 30# over normal because my suit and helmet were soaked. That was in an open cockpit DSR.
How protected from moisture intrusion is this unit, connectors, etc.?
Which puts the system in conflict with the GCR which says the yellow is in effect until past the incident.Quote:
Originally Posted by BeerBudgetRacing
The body of the unit is weather proof but connectors on the back are not.Quote:
Originally Posted by Farrout48
Brian
1) With the system currently being implemented by other sanction bodies, I would assume a regime for yellow incidents is pretty well hammered out. Probably a GCR change will be required.Quote:
Originally Posted by Rick Kirchner
2) (opinion) The yellow on 'your' display will be triggered at the flag station/signal board's assigned GPS location. These locations have to be preassigned. So, you will still be racing to the incident flag station/signal board (showing yellow) I would just not expect a yellow dash display until you reach the station.
3) Clearing the yellow/incident is a more challenging issue for the system. IF the system was able to know the GPS location of the car causing incident then you could turn the yellow off as soon as you cleared that GPS location. Probably not how it is done. I would assume the GCR is going to change to state no racing until the next station. That creates some issues: Should the signal board at the next station show green or no color/off? If there is a flagger, does he wave a green flag or not?
A lot of interesting small details to be ironed out.
Brian
Update: After watching the video below noted some interesting things:
a) The yellow 'display' came on way before the flag station associated with the incident. Could not make out a flag or signal board.
b) Green 'display' come on just after clearing the disabled car, way before the next flag station.
c) After virtual SC was cancelled the green came out just after passing a double yellow flag station. That could just have been a timing coincidence.
The one thing I took away from the video is that your race is being control by the 'display' more than the flag stations.
Clearly, not many people on this thread have watched the podcast with James Ballenger, the creator of Flagtronics, and Runoffs champion Todd Burris.
Here is a link to the broadcast. https://www.youtube.com/live/Ruhu1nU...ts9zkhjl6PFPEi
Many questions that have been asked in this thread are answered in this video. Pay attention to the on board vedeos.
Will this system replace flag marshals, will there be a subscription fee, will you get more green flag racing instead of following a safety car, will emergency vehicles respond to incidents quicker.
The Champcar series has required Flagtronics the past 2 years. Champcar owns a system, they set it up at each track.
Here is a link to the 2023 Champcar schedule.
https://racingcalendar.net/champions...ce-series/2023
Yes, Flagtronics works at Road America, Road Atlanta, VIR, Daytona, Sebring, etc.
I worked as a flag marshal at the Daytona Champcar event, on turn 5. Having never worked a Champcar I was expecting to see a demolition derby. I was surprised to see the cleanest racing that I have seen in decades.
127 cars started the 14 hour Champcar Daytona event. Over 90 we're running at the end. There was never a double yellow, safety car period. There were some code 35 sessions, which were short. The race did end 20 minutes early after a car blew a motor entering T7, the bus stop, after which 7 cars wrecked when encountering an oiled surface in the braking zone.
Champcars Flagtronics system includes light boards, which were mounted at each flag station. The light board responded immediately to inputs from our turns Flagtronic control unit.
I just finished watching the video as well.. I'd say it was about an hour and a half WELL SPENT. I would also say that SCCA has made a HUGE mistake by NOT referencing this video in comms to all racers. James (developer) seems to have addressed all .. or at least ALMOST ALL of the concerns brought up on this thread. He still didn't answer several of my questions.. BUT he does indicate that the system WORKS and the video indicates the biggest answers to our SCCA RACE PROBLEMS.. in that it provides VSC for racers that eliminates .. or significantly reduces.. trudging around tracks under DY .. trying to catch that leader while the leader has NOT slowed .. lap after lap.. or BFA (with the clock still running) waiting for cleanup... and/or FAILURE TO CALL for RED or BFA by Race control because they can't get 'the data' of exactly how bad the incident might be in a reasonable timeframe. It also addresses (previously NOT MENTIONED AT ALL) that the tracks/regions also have a $$$ commitment to make this work effectively... AND that THIS SYSTEM has already been in service and used by Champ Car for more than 2 years.
All things that we did not know.
NOW.. that it's clear (at least to me) that this system *IS* a good thing for all of us.. all we have to do is figure out HOW to get it into our race cars .. and, of course, PAY FOR IT!
SCCA didn't even send out an email to let us all know that this discussion was available. Pretty disappointing considering the impact and 'appearance from nowhere' in the GCR.
Thank you, Dave for putting the link in this thread.
Steve
The only question is who is going to pay for the $2755 corner station activation devices? I would think that most tracks will be more than willing to step up, particularly if a number of clubs require Flagtronics.
Now If I can figure out how to take it apart and make all the multi pin connecting cables needed to operate it in a configuration that I can fit into my car having the display reasonably visible....
Call it a business opportunity.Quote:
Originally Posted by Steve Demeter
Maybe it is in this thread-I have a question:
1.) Which "system" overides the other? Flagtronics of the corner worker(s) throwing the flags.
2.) Which "system" do I pay attention to?
3.) If Flagtronics shows a yellow, but I do not see a yellow flag thrown by the corner worker-HAve I just violated the rules? Will I get black flagged for ignoring a "FlagTronics Yellow" versus a real flag?
I am sure there are answers and I am sure the SCCA will generate a methodology to deal with this. (How does ChampCAr handle this?) For me that is probably a more important issue than cost or mounting.
looking into a crystal ball I'd bet the formula car community will either have to use the sedan unit of buy CAN dashes.
Won't the 'CAN' dash need to have a built in program to create the FT like images on its display?Quote:
Originally Posted by Rick Kirchner
Brian
Just raced two days with Champcar at Mid Ohio using this system. I have raced with them and Flagtronics before. By and large it works well. Are there instances where it's not perfect, yes. One example where it causes a bit of confustion - racing in a pack of 4-5 cars, incident - car spinning/off at the keyhole as we were turning in / entering (pro course), "LY" (local yellow) came on just as we were starting to accelerate onto back straight - same incident? Additional incident further down track? Check flag station down back straight, no flag and just that quick - green flag on Flagtronics came on as we cleared the incident zone. Champ instructions are (before Flagtronics) - yellow flag at one station, no PUY until you see the next flag station with no flag.
Pulls were handled by Flagtonics going full course yellow, once everyone slowed ("75% of race speed"), Purple 35 came on and trucks dispatched. Once tow was completed, back tp full course yellow (speed up), then green flag. This stepped slow down / speed up made for a safe / smooth transition between scenarios.
I am NOT a fan of additional expense, I like prefer "live" flaggin and "live" reporting of incidents. Both live and electronic have limitations. I do see some of the benefits of the Flagtronics and its limits. There is also the need to update firmware. I noticed at registration a notice reminign all teams to be sure they had the latest firmware version installed in the device. Not sure what limitation / issues one might have had if not done.
Bottom line, I like some aspects, I would prefer not to have the hassle/expense of another "device" to maintain. I think the path is clear that this is how the future looks. "If" it is, one system that works across several groups if prefereble to having to buy maintian several different systems, but the "monopoly" concenr is troubling. Watching carefully how it develops.
I'm sure there will no doubt be collaboration - unless two manufacturers can't get along.Quote:
Originally Posted by Hardingfv32
Brian
I am no expert, but does this video provide any insight about integrating the FT200 to a CAN bus device?
https://www.youtube.com/live/tWuOaC2...s3HfrVw3oCNPCa
That is the harder way. If you set the FT200 to always transmit, there is already a protocol in RS3 that just uses the flagcode. You can then set a few alarms on that based on the number that the FT200 sends. I am happy to help anyone that wants to do this.Quote:
Originally Posted by Dave Welsh
So, alarm lights not a FT 'like' display image.Quote:
Originally Posted by romoman
Brian
The alarm can be lights or a number of different things. What the alarm allows depends on the model of the dash. While the color TFT displays can make the page specific color, obviously the LCD dashes can not do that.Quote:
Originally Posted by Hardingfv32
Just general knowledge:
Assuming a TFT display, is the FT CAN transmitting data the AIM display can directly process into a FT200 'like' Image?
Or does the AIM need a firmware update to show FT200 'like' images?
Thanks
Brian
I know this topic has been beaten up for a while now, and is still a hot topic, but after this weekend at Watkins Glen with SVRA, I have a couple more thoughts.
1- While I did not have a Flagtronics box in the car, they did run FT LED panels at the corner stations. Those are AMAZING, and almost impossible to miss... awesome in rain/spray. I could see the end of the back straight board before I was on to the back straight. Those really seem to possibly be enough on their own, especially in darker conditions. The distance those could be seen was impressive, so congrats to FT, SVRA & the Glen for implementing them.
2- It would be a bit of work (or not, what do I know), but it would be AMAZING if someone could figure out how to make a receiver unit that you could plug a standalone LED strip into, then mount it how & where you want. Dash, roll bar, mirrors, etc.
If people are already able to make flexible & cuttable LED strips for random home projects that can be controlled via a remote... SOMEONE should be able to code something similar for our use.
https://static.lepro.com/media/catal...d6f82ab04d.jpg
There is no need for FW change. The AiM has screens already that show debris, checker, and mechanical. Then there are all the standard flag colors. You can also put your own text on it to say what it is (i.e. local yellow, full course, etc).Quote:
Originally Posted by Hardingfv32
That was my point a few posts back but after spending more time watching their videos, it will take more than a simple strip light due to how they are mechanizing this.Quote:
Originally Posted by Matt Clark
Black is a problem, so they make black with alternating white/off with a character. Then there's the speeds that show up in the purple. And you still need a way to show course condition like red/yel, passing (blue only, no need for red stripe?), red cross, and meatball. That's why I think they'll just tell us to use the sedan unit or buy a dash that the FT can be integrated with.
From a human factors standpoint, if digital flags are the future, we just ought to replace black with purple and be done with it, and then clubs could agree on a standard like flashing purple for all in, steady purple for an agreed upon speed, maybe alternating red/purple to indicate a meatball or a chat with the steward.
I don't see the 'speed' as necessary for purple display in the car. The station panels should flash the speed.Quote:
Originally Posted by Rick Kirchner
The whole point of the driver facing in car system is 'attention' not 'total communication'.
If it's going to provide any data other than a color it's not going to work. I will not be putting my head down to read something.