Gravity Vehicle C

[Balsa Man]

Well, one thing I have been successful at is generating some vigorous and interesting discussion; that’s good, and fun.
What I seem to have not been so successful at is conveying clear understanding of a number of key points. Trying to bring some detail into the evolving discussions seems to have only muddied the waters. Ah, were we all able to sit down together in the same room and have a real-time conversation……

My throwing the issue, and follow-up idea for one way to address that issue, came from coming to understand over the season a few things that relate to the scoring system, the performance parameters that can be…..engineered, and the measurement of those performance parameters.

It started out with recognition by a number of folk some time back of the human error in timing issue; the possibility of electronic/photogate timing was brought up, and seen to be impractical, if not impossible. I realized at that point there was an option to measuring/scoring what the current time to distance factor measures and scores, and tossed it out there. Do I think the “issue” is a big one; something that really has to be fixed?, no, it doesn’t affect most competitors, only upper end, to some extent. Do I think/feel something has to be done about it, or that what I threw out is the right, best, or only way to go?, not at all. It’s been a fun event this year; I’m comfortable the powers that be will come up with rules for next year that will be a fun challenge

Key points in review:

1) There are inherent, insurmountable limitations on being able to accurately and precisely measure start to stop time. The only viable means for time measurement is, as is currently laid out in the rules, manual timing, multiple timers, averaging of results
2) The magnitude of the error bars for this measurement is at least on the order of…. 0.2 to 0.4 second (at 5 and 10m), and that’s roughly on the order of 10% of run time to those distances.
3) For two similar vehicles, where one is about 0.2 to 0.4 seconds faster (to 5 and 10m), a “speed difference” of ~10%, the difference in max run distance is 50% (30m vs 20m)
4) Both run time (to whatever distance) and max distance are measuring exactly (and only) the SAME….performance parameters; the interactive combination of velocity coming off the ramp, and the friction management that determines the rate at which horizontal velocity is degraded/lost. We’re not- and I’m certainly not talking about “changing the goal” from speed to distance; measuring speed to a distance and measuring max distance capability is MEASURING THE EXACT SAME THING- the same set of performance/engineering factors. It is just doing so in a different way.
5) IF you could accurately measure time to a distance, two vehicles with the same run time to a given d, would have equal max distance. With manual timing, you can’t accurately measure time (to a precision of less than a couple tenths of a second, ~10%). With a tape measure, you can precisely and accurately measure distance. With a metric tape measure, and let’s say even one with just cm markings, you can measure that 10m delta to a precision of 1 in 1,000; 0.1%
6) If, in scoring two vehicles, time measurement is 0.2 sec off (of the “real time” difference between the vehicles), that equates to 10 points (0.2 x 50); that equates to the points for being 1 cm off target- 10mm = 10 pts; hence, if the difference in distance off is less than 1cm, then the timing error….prevails, and the “wrong one” wins; if the timing error is off by more, the difference in distance off at which the timing error “flips” the ranking is greater
7) When we get into ranking/scoring “upper level” vehicles – the ones that are both fast and accurate- with the scoring values that are currently used for time and distance off the mark – the inherent error bars of the timing measurement unavoidably mean there is a very strong probability of incorrect ranking; an inherent “unfairness”; a wild card. Not having a wild card at play is a good and desirable thing.
8) Measuring max run distance (instead of run time) is a way to totally eliminate that wild card. Combining a max distance score, with an accuracy to a point score, would result in correct, objective ranking of vehicles for combined “speed” and accuracy. I haven’t been advocating “removing limits on distance”, simply presenting the case that measurement of max distance capability is a much better, fairer, more accurate/precise way of measuring “speed capability.” My assumption was scoring, like this year, combining, with weighting a speed performance and precision performance factor.

There is indeed a real, and perhaps, maybe even probably, insurmountable, obstacle to this alternative approach, though; the need to align (by manipulation of the rules; the car mass and ramp height) the max distance capability inside the acceptable minimum space venues would have to be able to provide. At some ramp height and max weight, there is no way that even the “best of the best” – those magic, high-hours, high dollar confections capable of playing for a Nationals win could get beyond a distance that all venues could provide. Just for discussion, were event supervisors to (have to) provide a 10cm- or even 5cm high ramp, you’d get more precise measurement/ranking of speed capability than with time to a distance, and not need more than 10m of “track.” I’m not thinking this “common” ramp would replace competitor-built ones; just be a common tool used for & by all to measure speed capability. Runs to the target point would be from competitor's ramps. Given the disparity range between “good” and “not so good” vehicles (with some having trouble getting to 10 m off a 1m ramp), bring the competitor-built ramp spec down into the range that would assure a top vehicle wouldn’t go more than …..10 or 20m doesn’t seem a reasonable idea. Would that be an unacceptable “burden” on event organizers? Certainly no more demanding than providing a laser shoot box for Optics, or a playing field for Robot Arm. Yet, I’m sure there are some who would argue it was too much to ask… and maybe they’re right.

I agree with nejamimb’s observation that video timing seems a potentially viable timing option for Nats.

As to using a max distance measurement of speed performance (rather than time) driving and/or creating a technology & how much can you spend on bearings war, NO, it wouldn’t; you’re not changing the design or performance basis; not changing WHAT you’re measuring, you’re just changing the way it’s measured. The better the bearings, the faster it will go at any given distance, and the longer it will roll.

As to multi-hundred dollar skateboard bearings providing an advantage (whether you’re measuring speed(time) OR max distance), their design parameters (specifically initial and rolling frictional torque) are much less optimized to the demands of a GV than other much cheaper choices. They’re designed for MUCH heavier loads. As long as we stay within the realm of ball bearings (vs, say jewel bearings, or fluid, or magnetic bearings), small, high-precision bearings, designed for loads on the order of a few kilos is the path to minimum friction. The bearings in Super Pulleys, btw, aren’t either, a) a magical answer, or b) particularly expensive. $25 for a SP; bearing cost is significantly less than ½ of that. Actually did a spin test comparison of the bearing out of one; what we ended up using was better, and <$10 ea. To the extent anyone believes this is an issue that needs to be managed, a simple return on investment control would be easy – just adjust the weighting of the speed performance score and precision performance score. If it’s not going to get you meaningful score improvement, there’s no incentive to throw away $$s.

There are clearly different perspectives and technical takes on the distance capability at whatever ramp height and vehicle weight, and on how far the “best of the best” might be able to push. That’s a whole separate issue, and I don’t think there’s a practical way to resolve & quantify it to the point that at some set ramp height and vehicle weight, and required minimum track length, there would be certainty nobody would be able to over-run it. While needing an airplane hanger is clearly not the case, my sense that the 30m at current ramp and weight specs we’re getting is pretty close to the performance limits achievable may, to some extent be optimistic. I have to say, though, I really do find it hard to believe that someone could get a 100gr vehicle off a 25cm high ramp to go over 20m

As to multi-hundred dollar measuring tools needed, and measurement to a mm needed, that’s a non-issue; $25 or less for a 25m tape; or w/ a 10m tape, you could lay out 10 & 20m lines, and measure from them. Just measuring to the cm precision gets you to better than 1% accuracy.

In the grander scheme of things, this is all small potatoes. It’ll be fun figuring out what to do with next year’s rules.

[nejanimb]

Good summary. Only brief consideration gives me one thought - I don't think they (speed vs. max distance) are necessarily measuring the exact same thing. I'm guessing that slowing, if left to go the full way, doesn't actually happen linearly. My guess is that the current speed method and only having to travel a short distance puts a higher value on the velocity leaving the ramp whereas, by comparison, friction management becomes more important for maximum distance runs. Another note is that this way, speed might need to be balanced against braking concerns - maybe if you go too fast, you can't brake as well as you might have otherwise. With maximum distance challenge, that tradeoff could go away.

Doesn't invalidate the idea that it might be an effective alternative method to challenge the engineering, but my initial impression is they they are not measuring exactly the same thing.

I liked Electric Vehicle, where the best cars would be going along really slowly and had a target run time. Made this question much easier, though there are still always some timing concerns. Mission had that too - going for the minute long run or whatever it was, and going over was a big difference from being under or right at, meant teams had to try to guess what the supervisors reaction times would be. Timing is always tough.

[chalker7]

Balsaman is laying out a very good theoretical summary of how to separate the top handful of teams in a vehicle event. The problem is that we can't just deal with theory when writing the rules, we have to take practical considerations into account and think about all venues and teams. For example, if we limited the ramp height to 10 cm, some teams might not be able to travel more than a few feet, which would neither be particularly exciting to spectators nor feel like much of an achievement for the students (this is an important factor.)
When going through the rules, I actually think less about the elite teams than I do about the regionals-only teams. That's not SO policy or anything, but I want to make sure that events are approachable by all teams and that all tournaments can run the events. I've been discovering over the past couple years that far too many tournaments eliminate events from their schedules due to the perceived difficulty of either participating or running the event. My goal is to increase participation, not restrict it. Now, this all isn't to say that an event should favor random luck. Of course the best event is one in which the team that works the hardest and has the best solution wins while still allowing the team that starts a week before the competition to participate. Is that an impossible combination? Perhaps, but we're working on it and the important thing to understand is that there are a huge number of considerations that come into play when making the rules, some less obvious than others.

[Balsa Man]

Just 2 quick things

I'm guessing that slowing, if left to go the full way, doesn't actually happen linearly.

Correct, v vs t is NOT linear, at any point after leaving the ramp. When v is higher, it is closer, but the plot is still a curve.

Chalker7; hear, understand, recognize, and agree w/ everything you're saying. Recognize the challenge of of rules across a broad range of competitors that are fair to all. Over the years, the S-O organization has done a pretty good job with it. Appreciate all of you in the middle of it. Keep up the good work.
Yeah, a mostly theoretical discussion; trying to highlight a few things/relationships/factors that are not necessarily intuitive. Like early on in the season when there was a a lot of talk about a brachistochrone curve shaped ramp would produce higher exit v, which is dead wrong. A little pot stirring, because having to think about things you think you already understand often leads to new understanding.

[iwonder]

Correct, v vs t is NOT linear, at any point after leaving the ramp. When v is higher, it is closer, but the plot is still a curve.

This may be a stupid question... but wouldn't v vs t be linear because of the constant deceleration? I don't see how the forces acting on the vehicle change with the distance it's traveled to produce a curved v vs t... certainly a curved x vs t however.

[Jdogg]

Correct, v vs t is NOT linear, at any point after leaving the ramp. When v is higher, it is closer, but the plot is still a curve.

This may be a stupid question... but wouldn't v vs t be linear because of the constant deceleration? I don't see how the forces acting on the vehicle change with the distance it's traveled to produce a curved v vs t... certainly a curved x vs t however.

That's not a stupid question, since your right

[buzzbuzz]

Does anyone know the distance from nationals and any approximate times/distance scores that placed well?

[Schrodingerscat]

Correct, v vs t is NOT linear, at any point after leaving the ramp. When v is higher, it is closer, but the plot is still a curve.

This may be a stupid question... but wouldn't v vs t be linear because of the constant deceleration? I don't see how the forces acting on the vehicle change with the distance it's traveled to produce a curved v vs t... certainly a curved x vs t however.

That's not a stupid question, since your right

There will be some non linearity however, as for example air resistance is proportional to the square of velocity.

[fishman100]

Does anyone know the distance from nationals and any approximate times/distance scores that placed well?

Distance was around 916 cm from what I heard; perhaps bear will clarify this?
I was there during the last session and saw Grand Haven and Solon's vehicles; both impressive. I think Grand Haven must have hit the Target Point on their 2nd run since I saw one of the team members do a fist pump or something when he was retrieving the vehicle.
Did anyone see Rustin's runs? I was at an invite with them earlier this year, and even then they had a really great vehicle. (Saw their ramp at nats, and from what I could tell they didn't change their design.)

[GoldenKnight1]

Does anyone know the distance from nationals and any approximate times/distance scores that placed well?

Distance was around 916 cm from what I heard; perhaps bear will clarify this?
I was there during the last session and saw Grand Haven and Solon's vehicles; both impressive. I think Grand Haven must have hit the Target Point on their 2nd run since I saw one of the team members do a fist pump or something when he was retrieving the vehicle.
Did anyone see Rustin's runs? I was at an invite with them earlier this year, and even then they had a really great vehicle. (Saw their ramp at nats, and from what I could tell they didn't change their design.)

1st place score was very close to 160 points. About 3 seconds and less than a cm.