ZO GPS Accuracy

[gavski]

Quick question. How accurate is the GPS position fix that ZO uses?

 

The United States government currently claims 4 meter RMS (7.8 meter 95% Confidence Interval) horizontal accuracy for civilian (SPS) GPS. Vertical accuracy is worse. Mind you, that's the minimum. Some devices/locations reliably (95% of the time or better) can get 3 meter accuracy.

 

In aviation, we use a ‘blended’ position solution to the problem of accuracy..so the question is why doesn’t ZO? Why isn’t there a sub buoy magnet at each end of the course with a pickup in the boat that the ZO can get a ‘real’ fix on before it enters the course?

[Tom351]

I think this is the GPS receiver that both PP and ZO use:

https://static.garmincdn.com/pumac/GPS_18x_Tech_Specs.pdf

Accuracy info is on page10

Not sure if it uses GPS or WAAS though.

[eleeski]

ZO beeps at every buoy. The timing of that beep is quite accurate and consistent. I was skeptical of end course times so we checked against hand timing - ZO was right on at every buoy. The pull feels consistent. ZO is doing something right with just the satellites.

 

Waterskiing is not a critical activity like flying. There is the beep cross check. GPS drift at a specific location is relatively slow so if you calibrate your course it will be good all day.

 

Eric

[gavski]

Standard GPS accuracy is 15meters...the WAAS - The Wide Area Augmentation System is an air navigation aid developed to augment the Global Positioning System. It is 3 meters...10feet....

 

Any numbers of factors can degrade this accuracy further...even so, when working well, the ZO can be as much as 10ft (3meters) out....this would account for many of the variables as to the different ‘feel’ of the pull behind the boat...

 

A simple ‘anchoring’ position fix from a magnetic sub buoy would tell the ZO exactly where it is at the start of each end of the course..

 

A 10feet tolerance is accurate enough to auto land a Boeing 777 onto at 180ft wide runway...compare that to a 10ft tolerance for ZO system and a slalom course....

[eleeski]

Surveyors are able to get centimeter precision with satellites. An aircraft has quite different challenges than a surveyor or a boat that never leaves a few hundred meter radius.

 

ZO works. It does so without extra hardware and complexity. We don't need to solve a problem that isn't there.

 

Eric

[Bruce_Butterfield]

You are confusing accuracy with consistency. Accuracy means that when ZO hits the gates, the satellites report that it is within x meters of the actual lat/Lon position.

What we care about is the consistency that once we map the course, each time the boat hits the gates it is within a small fraction of a meter of the mapped position. True positional accuracy is irrelevant.

[gavski]

That’s the point..even when you map the course, your are mapping it to a 3meter tolerance..therefore every time you come down the course, the boat will be somewhere within 3meters and be different each time...the timings are bang on because the ZO algorithm will tell you that...ZO will give you acurate speed across the water same as your sat nag...the problem is that you are overlaying a virtual slalom course and it will never be more accurate than within a 3 meter radius or the actual slalom course..

 

@eleeski, how can you say that when we are in a sport that demands courses to be surveyed, boats speeds measured, boat path monitored, ropes measured etc etc all to mm accuracy...geo surveying uses triangulation etc to narrow the positional error - and is static....

[MISkier]

@gavski, I don’t believe that the ZO pull incorporates any pull characteristics relative to the buoy location. It just responds to your pull and returns the boat to speed.

 

There is one exception: the higher speed before the entrance gate. That could be slightly affected by thinking that it should shut that off sooner/later than the actual gate, but I think the risk here is negligible.

 

I don’t see a need for magnets again. I’ve never noticed ZO doing something to me other than reacting to my pull on the boat. I’ve never felt it suddenly speed up or slow down in my glide because it thought the course was someplace else.

[Tom351]

One other thought: 3 meters seems like a fairly large tolerance for location, but the GPS receiver spec sheet lists the speed accuracy at 0.1 knot. So while you might not want to place buoys based on GPS readings, it's accuracy for determining moving speed seems pretty good. I assume this is because it takes 5 readings per second, and the 3m margin of error gets quickly averaged out across the multiple readings and the speed calculation derived from them.

[eleeski]

If it beeps at the buoy, it's right on. ZO reacts to the accelerometers for speed control - not position. There is a vague speed target relative to position (overspeed anticipation of the gates) but once the skier pull affects the accelerometers the positioning is irrelevant.

 

Timing is relative to the beep and GPS does not vary by much over the length of the course or day - regardless of the accuracy. The times are real and well within tolerance.

 

The real world complaint with ZO is that its times are too perfect. If you pull hard, ZO hammers you to keep the time perfect. Not enough give to match the feel of hand driving or PP Classic (that it replaced).

 

You are focused on a variable that doesn't matter. It has been successfully engineered out of ZO. My ZO beeps at the right time and consistently feels the same and has for years. No problems and the transition to tournaments is perfect. ZO rocks!

 

Eric

[gavski]

ZO will always give you actuals...the accelerometers are very accurate, and the system is very clever. I was just suggesting that the ZO position be zeroed every time that it enters the course by using a fixed anchor point - like the Right hand gate buoy and not a guessed/ relative position based on a gps location that is subject to a 10ft tolerance...once zeroed on a known fixed point, then let ZO do what it does best...

 

I bet that even after mapping the course, it won’t take long before the beeps start to wander slightly from the buoy location..I have seen this happen..even by a couple of feet, it is still a large error...despite this, the timings are spot on...funny that, the ZO will still give actuals even when there is a position shift from where it thinks it is and where is actually is....

[MISkier]

There were previous side-by-side tests (in the same boat) with ZO and PerfectPass (timing) and they matched. I can’t remember if they used magnets for the PerfectPass, but, if they did, then the physical timing should match the virtual.

[Jody_Seal]

 

Speed Control Parameter :Slalom and Jump:

Recommended remapping Before event and as necessary if off more than 2.5 meters.

 

Something must be amiss as this is from the recent rules addendum.

[gavski]

it really doesn't take long for the gps position to wander off from the mapped position...in a sport where lateral deviation by the boat by a couple of inches can be the difference between a world record or not, ±2.5m seems like a ridiculous tolerance to accept..especially when it's random and ever changing..

 

ZO uses a virtual map of the course that it bases its timings from, so as long as it's holding the speed to ±0.1kt then the time will ALWAYS be accurate and the same as PP, even if the position has wandered off. The ZO can't tell if it's 'known' position is accurate...everything is based around a relative/ guessed position and a virtual course.

 

if the ZO is 3ft behind itself down the course, how do you think this will impact the pull?? what about from the other end when it is 3ft ahead of itself?

 

next time you drive down the course, take a note of where the beeps are happening..then compare from the other end...

[BRY]

@gavski I believe you are tilting at windmills a bit. I think this is from a misunderstanding of what ZO is and does.

ZO is a speed control system and only a speed control system. It is not an autopilot or precision location device. For instance your analogy to a airplane, "10 feet tolerance is accurate enough to auto land a Boeing 777", makes no sense. ZO has no way to position the boat, to hit a course (or a runway). The driver does that. If the driver is left, right or center ZO will give a very precise speed. ZO's ability to maintain a very precise and correct speed has been proven over and over, ad nauseam. ZO is speed control only.

 

As ZO is a speed control device only, the ZO pull does not incorporate any pull characteristics relative to the buoy location. ZO reacts to accelerometers and GPS differential for speed control - not position, as @eleeski and @MISkier said. It does not " make up time", it does not have to. It's ability to respond and keep speed is so good it does not have to (be that good or bad).

Side note: In tournaments I get 95's behind 6.0 and 6.2 liter engines always. Behind the 5.3's sometimes 96 & 97's. Behind 5.7's I will get 98's now and again but if I am really on I get 95's with them. It could easily make time and always give 95's every time but does not.

 

The only time location is used is for the overspeed into the gates. But once the skier pull affects the accelerometers the positioning is irrelevant as ZO does its thing to set speed. ZO does not need to overspeed into the gates, it can hold flat speed in easily, that was programmed in. That was done as Classic PP, being rpm based, runs fast with no load. The overspeed to gates was put in then to mimic PP as people were used to that. As said above variation for this overspeed location just does not matter.

 

In a sense "timing" with ZO is not really necessary. ZO has been tested and investigated time and time again over the years, ad nauseam. Go back every couple years on this site and you will find massive threads with people new to ZO attacking it. Every time it comes up to work, exactly as it says it does. There are those that don't like what it does but that is something entirely different.

 

Specific to your question: "if the ZO is 3ft behind itself down the course, how do you think this will impact the pull"? Answer: not at all, location has nothing to do with the pull.

 

 

 

 

[gavski]

Noted... so what is the point of timing if it is that irrelevant to the pull/ course??

All I was stating is that the ZO position that it uses is only as accurate as the GPS fix, which is at best 10ft....

If there was a system to zero the virtual ZO position to the actual position everytine the boat entered the course, then the timing would mean something...at the moment, it is referenced to a guessed position...

 

As stated, ZO will give you actual timings regardless of the position error... so what is the point of the timing? I think you will find that there is a relavance to the speed/ pull..it has been shown to me how if ZO is slow at the first segment, it will nail you in the second in order to stay within time/ speed tolerance...

 

[gavski]

I am not a ZO hater.. just trying to find a way to improve the system.. how long did it take to get the + feature integrated??

 

The positional accuracy has always been an issue for me.. v easy to fix. The timings are based on a known distance...all useless if the position info is junk...

[gavski]

Yup...posts got crossed...still can’t believe that a 2.5m positional error is acceptable in order for the ZO to base its calculations on...

 

@Bruce_Butterfield talks about consistency..how is that possible when the ZO position is out? Position shift between the satellites and ZO is not linear...it depends on numbers of satellites, position of satellites, quality of signal, almanac date, jamming, military interference etc etc...so bang goes the constancy..every pass is different..

[rfa]

@gavski, I believe I understand your point. Since it is possible to get a fixed signal to identify/locate the course, why rely on a GPS signal that may drift over time? Fair enough. However, as @BRY explained, ZO is a speed control system. It will pull you at the set speed anytime it is ON, whether in the course or free skiing.

"The timings are based on a known distance...all useless if the position info is junk..."...

No. The timings are based on a known speed.

[gavski]

@rfa timings are based on distance/speed...I assumed the speed was a given in this case...

[Bruce_Butterfield]

@gavski you are still not understanding.

 

For sake of illustration let’s say the ZO gps positional accuracy is 10m. That means when you go through the gates, the ZO absolute position could be off by 10m from a dead nuts survey.

 

However neither the skiers nor ZO care about absolute positional accuracy. The important point (mainly for the gate over speed) is that when you map the system and tell it where the gates are, every time after that it will be within inches of the mapped position, drifting slightly over weeks and months. That’s relative accuracy, or what I initially called consistency.

 

If your concern was an issue, every time the boat went through the gates, and the beep at every buoy, could be off in random amounts by the gps accuracy of 10m. So you tell me: when you drive through the course are the beeps ever off by anything approaching 10m in random amounts? Or are the beeps at the exact same point (as close as a human can judge)?

[eleeski]

When I fly my Cessna from San Diego to Palo Alto, the 3 meter GPS accuracy is incredible. But that's not how ZO works.

 

ZO might not actually measure distance to calculate times. Speed variations can generate the times. However it figures the time, the times agree with different measurement techniques whenever cross checked.

 

We used to look over the side and hit a stopwatch when we crossed a buoy to measure times. Not off by 3 meters but not very accurate. Comparing the beep to crossing the buoy is a common skill. Believe us when we say ZO times and pull profiles are accurate.

 

@gavski is probably @MS under a different name bashing ZO from a different angle. Some of us enjoy the predictable, easy and consistent pull from ZO. Complain about the cost of ZO boats and you'll get some support. But technically ZO rocks.

 

Eric

[rfa]

@gavski, yes the speed is a given and the "trusted"/accurate parameter in ZO. The beeps represent traveled pre-set distances (standard course dimensions!) based on boat speed.

[Zman]

@Bruce_Butterfield is right on. So long as your beeps occur within a tolerance of the gates and boat guides, all is good. The beeps might drift over weeks or months, but on any given day the beeps will be almost exactly the same every pass. Maybe a little early going in one direction, and about equally late in the other direction.

@Jody_Seal noted the rules allow +/-2.5 meters. @Horton posted that portion of the rules in another ZO thread the other day.

I suspect most of us remap well before we see 2.5 meter drift.

[BRY]

@gavski you say "it has been shown to me how if ZO is slow at the first segment, it will nail you in the second in order to stay within time/ speed tolerance..."

What do you mean by segments? How far off? Sure your not thinking of PP?

PP talks about segments as first (1,2,3) and second (4,5,6). PP will +/- second segment (buoys 4,5,6) to try to get in tolerance for overall course (classic and Stargazer, don't know on ZBox). Getting the front and back segments to balance when I put Stargazer in my old boat was a bitch, took a while. Once I did it was a nice ride.

ZO is ABT only so usually talk ball times, not segments. So are you meaning one ball time is slow so speed from 1 to 2 increased and the all other balls on? I have seen segments off by a hundredth here and there but no consistent pattern of slow/fast or fast/slow. ZO splits vary incredibly little. Its the power coming on people feel, keeping speed, that seems to "hammer" them, not a change in speed. It is a result of "hammering" the boat first, at 220 it is something I am intimately familiar.

[BRY]

ZO is based on GPS. GPS works on signal time shifts. For speed/distance the location is not needed as shift is what is measured, from one fraction of a second to another, from many satellites. Very accurate.

Where it thinks it is may be 10ft (or whatever) from where it really is, but move it and it knows very precisely how far and how fast it moved relative to the last measurement.

 

Want more than that, Google GPS. There are some great papers that explain it all.

Signing out, drivers home, gonna to test the ZO in my boat about 10 passes worth.

[HighAltitude]

I'm not sure 3m accuracy is actually true. It might be significantly better. GPS accuracy is predicated on calculating the x,y, and z axis. If the algorithm that ZeroOff uses keeps the altitude constant, I would think the accuracy would go significantly up. Just thinking back to my GPS fundamentals class I took about 15 years ago.

[Intheday]

Working in agriculture tested GPS is within 20cm direct from satellite. With a ground station within 2.5cm. So as long as you calibrated Z-O is likely within 20cm so I will take within a handspan.

 

 

[gavski]

so why bother with mapping the course?

 

what's the point if ZO will always give you actuals? what's the point if the timing is apparently meaningless? gross-error check? has it ever been wrong?

 

if this is truly the case, then i accept what has been said, and i thank you all for educating me.

 

GPS accuracy is a mute point...you may get better than 3m, but highly unlikely and certainly not consistent as that is the preserve of the military...the accuracy that we all enjoy from GPS satellites has been deliberately degraded for obvious reasons. no problem in civil aviation and car satnavs - not so good for projectiles aiming to fly through a window sized hole in a building hundreds of miles away...even then, to get pinpoint accuracy to levels we are used to in this sport would require laser guidance from someone looking at the target...

[DUSkier]

Timing is also used as a double check to make sure speed has been set correctly. I have had bad times with ZO, VERY rare and there have been external factors e.g. hardware faults, weather? loosing Satellites at a certain site next to a cliff.

As a side note I do not believe I have ever asked a driver if mapping was good feeling something weird at the gate, I would notice more so what a hand was doing on the wheel =)

[Drago]

The beeps and times are for the boat crew's entertainment only. I trust my sources on this (3 pandas told me)

[Roger]

@gavski - "what's the point if ZO will always give you actuals? what's the point if the timing is apparently meaningless? gross-error check? has it ever been wrong?" - ZO gives the the actual time it took travel the distance, it DOES NOT always give actuals. For example, the first Big Dawg held at Denver (high altitude which robs engines of power), ZO was returning times as slow as 17:05 (and a few even worse if I recall correctly). With enough power, times will usually be within .01 or .02, however I saw a time of 16.99 during a tournament at Palm Bay (Sea level) this year while driving a 6.2 2019 Nautique.

[DW]

Just an FYI to earlier comments on why have timing or timing data displayed and sorry if I misread your comments - I have had out of tolerance times a few times behind a ZO boat (slow, never fast). No idea why it happened and it never happened when I drive so I have no troubleshooting insight on it so I chalk it up to OE.

[LeonL]

Why display times? Only because the rules say we have to (tournament situations). There have been instances where the driver failed to set the correct speed and this was evidenced only due to noting the time. I was boat judge once and on a skkers's first pass (asked for 34) at the end the driver said "93". I thought that a bit odd, so I asked "what 93?" He responded 17.93. Error revealed.

[Zman]

@Drago Are you also saying that the ABT times that coincide with the beeps are just for boat crew entertainment?

Or, am I wrong in thinking that the ABT times coincide with the beeps?

[ForrestGump]

Keep in mind that while GPS only has a published accuracy of 2.5-3m, companies are layering proprietary algorithms on top of the standalone base signal to increase the accuracy significantly. Which is how the previously mentioned agricultural gps gets down to 6 inch accuracy. Since you don't know what algorithms that ZO uses(we know they use some sort of proprietary algorithm, though) you don't actually know what the true accuracy is. I'd venture to say sub 1 meter corrected accuracy.

[Tom351]

Great point by @BoneHead - and on top of this, the latest ZO systems use dual receivers- which I would imagine only increases the overall accuracy beyond the published spec of either receiver used individually (not that location accuracy is all that important in a speed control system anyway)

[ForrestGump]

What must be kept in mind is that there's far more to any GPS system, than just the receiver and its inherent standalone SPS signal accuracy.

 

The "Standard DGPS" belongs to the first generation of equipment used in large volume. It has an update rate of once per 30 sec per satellite. This was available in the mid 1980's and achieved 3 to 4 m accuracy. It is the original US Coast Guard design.

 

The next level, labeled second generation, began coming on line about 1990. This generation achieved "sub-meter" accuracy, about 70 to 100 cm. It uses about a 5 - 10 sec update cycle.

 

Using narrow correlator techniques, and the civilian SPS signals, 20 cm DGPS has been demonstrated using only ranging signals. There are now several manufacturers achieving 10 cm accuracy levels.

[Drago]

@Zman , I'm saying the boat averages damn near the speed you tell it to go and the beeps are where an extremely accurate course would have its buoys. It's just redundancy so we feel good, or as @LeonL said above.

The buoys aren't for timing anymore, they are so we go straight, and can score the skier.