AUTOPILOT = PERFECT BOAT PATH???

[BraceMaker]

@The_MS Sure but lets set it up for a 160 lb 36 mph 41 off skier.

 

With ZO I still think they've just failed to have a meaningful calibration mode - regardless of the boat/horsepower/prop/elevation. The way to do this is to have a mode where the boat cal mode is engaged and you drive down the lake as per usual. Only first the boat goes to 15mph and settles in then it applies throttle and measures how much response the hull is giving. It repeats until it has a setting that matches what is defined as the amount of acceleration. Then it goes on to 18mph and so on and so forth.

 

Now if you go out and calibrate a 2008 ps 197 or a 2020 MC 6.0 liter it should feel the same to the skier. Different prop? Recalibrate. feel like the engine isn't pulling like it should? Recal. At the end of the cal it should be able to confim exactly how much juice it gives.

[BrennanKMN]

@BraceMaker What is your definition of perfect? There is no such thing as perfect.

 

Even if you have a sled under water with 10,000 HP pulling it and more steel than the empire state building supporting it, no matter what when you add load it slows down. That slow down will just be so minor it is imperceptible.

 

Having that deviation will be exactly what makes it not like a brick wall to ski against. Just like @ral mentioned about the early versions of ZO, nobody liked the perfect rock solid exactly 36 MPH pull. That is why we have the ABC123 now among other tweaks.

 

Why not start with keeping the boat path inside of the allowable tolerances and you can adjust from there? You can always tighten up the response time to limit deviation. Like I mentioned earlier, with the proper equipment you could have a deviation of less than 0.5mm or something. Is that too much of a deviation that you'd notice?

 

Even if you have a 'learning' system, people are unpredictable. Your pullout might be normally 300lbs of force, the next day it is 200lbs because your tired. The system has learned that you usually pull out with 300lbs of force so it anticipates that and corrects for it. You have just created a center line deviation that created a disadvantage to the skier.

[BraceMaker]

@BrennanKMN I think we agree about deviation I think we disagree about reaction time.

 

My comments are informed by discussions like https://www.ballofspray.com/forum#/discussion/22571/boat-driver-effects-on-feel-to-skier/p1

 

If Chad Scott can be referenced as source material I am stating that a closed feedback loop controlling for deviation is by definition reactionary. It's a driver waiting to feel you hook up instead a driver who knows you are about to and would feel yucky to nearly anyone and specifically to the fact that it's tech for top end skiing (like record scores being dumped) when it matters.

 

A closed loop reaction system would be fine for public courses on open water or very casual course skiing.

 

[DanE]

@BrennanKMN A perfect calibration mode for ZO would be a standard that makes the pull identical behind every hull/engine/transmission/prop combination possible.

[rico]

I don't see why you couldn't get a system that anticipate as well as react to the input form the skier.

[BrennanKMN]

I guess I fail to understand why reaction time really matters if you can correct for a deviation with a very tight degree of accuracy and in fractions of a second.

 

Would a skier notice the difference between a driver anticipating a pull and correcting the boat bath slightly before or as the pull came or a boat that never left center line more than 2cm regardless of when the pull occurred?

 

To me a 'learning' system goes right back to the root of the problem. How are you going to bring your learned settings with you? Is each skier going to carry a flash drive with their settings and hand it to the driver before each set?

With a system that isn't learned it makes the boat path very consistent between different boats, drivers, settings, lakes, wherever.

 

Regardless, it would be a super fun learning experience to try skiing behind a perfectly straight pull. I think a lot of us would find it quite eye opening.

[DanE]

@rico My guess would be budget.

We are talking about a fraction of the market for three event tugs.

[ral]

@BraceMaker, a Control System can react/preadapt to many variables, including position in the course, data from earlier in the course, etc... and can consider many more variables much faster than any human being. Modern fighter jets FBW systems can adapt extremely fast to changes in areodynamics/loss of aileron pieces/etc..., and make flyable supermaneuverable jets that would not be able to be controlled by a human.

 

AI gets better trained in these type of problems by simulators/digital twins, rather than manual input of skier profiles.

[ForrestGump]

And yet Tesla's seem to keep running into other cars while the autopilot is on.

[addkerr]

@ForrestGump So do regular non autopilot cars, as well as non auto steer boats running into stuff. People are stupid, computers aren't perfect but at least you can adapt and improve a computer, unlike a lot of people I know.

[eleeski]

Click a stopwatch as fast as you can. .15 second is my human best. Computers have millisecond reaction time. This could allow the reactive response of a computer to match the proactive response of a human.

 

The allowed tolerances and historical use of them probably doesn't require "learning" to stay within those tolerances. Chasing some holy grail of 2cm variation is difficult to build, possibly quite different in feel, and excessively expensive. And not necessary.

 

Care must be taken to not use the millisecond response time to make a pull that is not humanly possible - even if it is more accurate.

 

Anti weaving programmed in is as much against the rules as weaving intentionally. Allow as much use of the tolerance as needed to avoid cheating the skiers by weaving against them.

 

Auto steering is inevitable. It's possible and will prevent another French scandal.

 

Eric

[liquid d]

Driving is also part of the fun of going skiing. I would never be interested in a system that would take that away.

[teammalibu]

@Zman Wat cha talkin bout Willis, its extreme shuffle board or nothing!

 

[rico]

@BrennanKMN I can assure you that if the wheel is not already turned the right way in anticipation for the skier’s pull, the boat is going to move, no matter how fast a system responds

 

@liquid d i agree. I enjoy driving and being in sync with the skier.

[BrennanKMN]

@rico I agree 100%.

The same is true about your statement, if you are turning the wheel in anticipation you are potentially causing the boat to move prematurely as well.

 

My point being if you design a system that reacts to a deviation in 0.01 seconds, does that movement really matter?

[thager]

Only if the bear craps in the woods without the tree hitting him quietly.

[Horton]

@BrennanKMN elite drivers do anticipate. You can not just wait to feel the boat move. On the other hand a steering system powered by SkyNet will likely move fast enough to only be reactive

[jayski]

@Horton I agree and disagree, with the billions of calculations per second I am sure they can make a program to anticipate fairly close how much a skier is going to load and when off the next buoy based on the load, duration, and release into that turn, with of course the help of banked memory from having to enter skiers name into system so it can track the 'habits' of the skier. I bet it could (if you threw all the cards at it) be an amazingly awesome pull...I am sure ZO off could feel a million times better if it tracked skiers 'habits'. probably wouldn't need a/b/c/1/2/3

[Chad_Scott]

and people complain today about the cost of these ski boats

[eleeski]

Number analysis: 16.08 seconds for the course. Entry and exit gates take a bit of time. So maybe 14 seconds dedicated to the buoys. 14/6 is a little over 2 seconds per buoy. Not too much steering response for the time behind the boat or when there isn't much load on the rope. So maybe half a second where the skier is really pulling the boat off line. With human reaction times around .15 second, a human driver will dally for a third of the critical time - leading to bad boat paths and feel. The human driver that anticipates accurately will time reactions properly for both the skier and path without weaving either way too much. A computer with millisecond response time can react quickly enough to match the skilled human.

 

@rico Yes the boat will move. It does now. Keeping the movement in tolerance (like for skiers in the past) is the goal. Zero movement, especially if it complicates things, is not necessary. If zero tolerance requires weaving against the skier, that's not proper as intentional movement against the skier is not allowed by rule (obviously, weaving against the skier involves movement - even if the boat ends up perfectly centered at the buoy).

 

A computer control will have the same forays into the tolerance for everyone. As long as the program stays in tolerance, the playing field will be level even if the path varies from exact.

 

@Chad_Scott I agree that boats are ridiculously expensive. But ZO is a cool feature and worth the $1800. A Perfect Path might also have the same value (and cost?). Losing all those French scores represents a huge and expensive loss - way more than a cool feature might cost.

 

Eric

[Chad_Scott]

Boats already exceeding 100k. And we keep adding cool features. They don’t get any easier to buy or sell is my point. Plenty of drivers that can keep it in the middle.

[Horton]

@Chad_Scott Shut up. You just don't want to lose your gig to a robot.

[Chad_Scott]

I’m not a driver. I’m a skier. I don’t want to be unable to sell these boats people complain about the price of now. And for the record. An automated boat driving system will likely feel like a steel cable down the lake

[ALPJr]

Circa ‘75 tech - the driver jams a short paddle between the driver’s seat and the steering wheel, sits in between the split windshield facing the transom, hits it, then scooches way up onto the front of the bow of our ‘69 V-143 with a 50hp to get the bigger guys up on a slalom ski. The tracking was great.

 

 

[swc5150]

Check out the new Nautique Paragon, and you'll feel much better about boat prices in our sport. The sales tax on that tub is more than our first MC cost new.