Ball speed testing after matches

[Ken Loyd]

At the Portland Regional on more than one occasion the head referee announced that a robot was going to have the ball speed checked after a match had been completed. Did this happen at other regionals? I never heard if the speed was excessive or if a score was changed.

Ken

PS Steve and Glenn, we stopped by the 254 pit and were told you are not working with them this year. I hope you are both involved with other teams. FIRST needs both of you.

[Conor Ryan]

At New Jersey this happened on 3 occasions that I can remember, 25, 293, 303 all were tested on Friday some time. 25 and 303 both were shooting around 11.7 m/s and 293 set the new land speed record at a whooping 15.7 m/s which they later turned down to about 11.5 m/s

The test unit consists of 2 light sensors that are spaced 1 meter apart and are connected to a computer where the time intervals were captured and then the speed calculated. The unit itself is then held right at the exit point of the ball. The only teams that tested were the ones that the ref's felt may be shooting at an excessive speed.

[Corey Balint]

At New Jersey this happened on 3 occasions that I can remember, 25, 293, 303 all were tested on Friday some time. 25 and 303 both were shooting around 11.7 m/s and 293 set the new land speed record at a whooping 15.7 m/s which they later turned down to about 11.5 m/s

The test unit consists of 2 light sensors that are spaced 1 meter apart and are connected to a computer where the time intervals were captured and then the speed calculated. The unit itself is then held right at the exit point of the ball. The only teams that tested were the ones that the ref's felt may be shooting at an excessive speed.
Those three tests you speak of happened late Thursday night. We werent in question, but just out of curiousity, we decided to check it. The head ref said we were never in question at any point. I know the bots they had checked after matches on thursday, three of them, were all clocked too fast, and all had to slow down their system.

[dangerousdave]

At the VCU Regional, I know team 435, Robodogs, were speed tested at least once. I stood 20 feet away as a tape measure was laid out and their robot was tilted forward at an angle and test fired. There did not appear to be any problems found. I believe it was done on day 2 of competition and the procedure took at least 20 minutes. Their robot design is excellent and a pleasure to watch. :)

[Kims Robot]

I have a really stupid question... and yes, I know some of this goes on GP, but how are they really going to know how fast a ball is going during a match if they test the robot after?

For example, our shooter at full speed can shoot nearly 65 feet. We are trimming it down in software so that it only goes the 12m/s or 35ish feet... but what prevents us or any other team with that capability from modifying our code or using some button to "show" the judges it only goes a certain distance? On top of that, our shooter has variable speed controlled by the driver, so our intention is to have max variation at 12m/s, but how are the refs going to know if we are at max speed or above?

My guess is the whole answer is GP... but I know a lot of teams, it might just be ignorance... many teams probably never really got to test the real speed of their shooter... distance would be the only predictor.

And dont get me wrong, we are going to do everything possible to make sure we are under the 12m/s, but it just seems if our team has the capability, what prevents any of us from doing it intentionally or unintentionally?

To me... other than for safety reasons, it seems like this is a really really difficult rule to enforce.

[A. Snodgrass]

I have a really stupid question... and yes, I know some of this goes on GP, but how are they really going to know how fast a ball is going during a match if they test the robot after?

For example, our shooter at full speed can shoot nearly 65 feet. We are trimming it down in software so that it only goes the 12m/s or 35ish feet... but what prevents us or any other team with that capability from modifying our code or using some button to "show" the judges it only goes a certain distance? On top of that, our shooter has variable speed controlled by the driver, so our intention is to have max variation at 12m/s, but how are the refs going to know if we are at max speed or above?

My guess is the whole answer is GP... but I know a lot of teams, it might just be ignorance... many teams probably never really got to test the real speed of their shooter... distance would be the only predictor.

And dont get me wrong, we are going to do everything possible to make sure we are under the 12m/s, but it just seems if our team has the capability, what prevents any of us from doing it intentionally or unintentionally?

To me... other than for safety reasons, it seems like this is a really really difficult rule to enforce.

It is. Its almost impossible to enforce, or it seemed like it at points. That might just be me though. How can one accurately tell from eyeballing if something is relatively close to 12m/s. Sure, you might be able to tell if something is going insanely fast, but its very difficult to actually say for certain that yes, that robot is violating the speed rule. And its possible that it will be left up to the decision of the referees whether to call somebody for violating the speed rule or not.

[Lil' Lavery]

Well, if your shooting it 65 fps, you'll be hitting shots from about 3/4 court, which might be a tip off :yikes: :p
But otherwise, yeah, it's an issure of GP and honesty.

[Greg Needel]

It is. Its almost impossible to enforce, or it seemed like it at points. That might just be me though. How can one accurately tell from eyeballing if something is relatively close to 12m/s. Sure, you might be able to tell if something is going insanely fast, but its very difficult to actually say for certain that yes, that robot is violating the speed rule. And its possible that it will be left up to the decision of the referees whether to call somebody for violating the speed rule or not.


well alot of this is all on GP but there are some ways to tell. The distance away from the goal, the height of the ball, the trajectory (flat or parabolic), etc. Once you have seen something shoot at proper speed it becomes instantly apparent if you are over speed.

[Zach Purser]

At the VCU Regional, I know team 435, Robodogs, were speed tested at least once. I stood 20 feet away as a tape measure was laid out and their robot was tilted forward at an angle and test fired. There did not appear to be any problems found. I believe it was done on day 2 of competition and the procedure took at least 20 minutes. Their robot design is excellent and a pleasure to watch. :)

Wow, I'm blushing, thanks for the compliment. Our robot was actually tested twice. The first time it was tested while sitting level and we passed the test, but then we were contested a second time because we shoot at a 60 degree angle rather than 30 degrees, so they placed us on a 30 degree slope to fire and we came in about a foot under the maximum distance. It was really more an optical illusion with our robot, because our flywheel maintained the momentum we could rapid fire a continuous stream of balls.

[Wyatt]

At the portland Regional, our robot was tested for 'shooting' velocity(we scored about 30 points in one dump, on the low goal). Turned out we rolled balls out at 1.2 m/s -_-

[Al Skierkiewicz]

A ball speedometer will be available at each regional in the future. It will take the guess work out of shooting for distance. As many teams have already figured out, back spin on these balls cause the trajectory to flatten out. If you are unsure of your ball speed and are prepared to trim the speed down to specification, you will have the chance with the speedometer.

[A. Snodgrass]

well alot of this is all on GP but there are some ways to tell. The distance away from the goal, the height of the ball, the trajectory (flat or parabolic), etc. Once you have seen something shoot at proper speed it becomes instantly apparent if you are over speed.
This is true enough, and it was obvious in some cases. But when the speed was close it was difficult to tell if they were a smidge over or not. And yes, in a lot of cases it was possible to tell when the robot was over. And some were caught because their shots were obviously travelling too far for the angle shot at etc. There was one at PNW that was only .6-.8 M over. And while it was possible to tell, I must admit it was much much harder to see.

[Joe Ross]

Somebody asked us to be tested after our first match. The funny thing is we had only passed the speed test right before that match. We shot 11.5 both times.

[Billfred]

I manned the speed trap at Florida for a few hours, testing the teams as the refs asked. Two teams that I tested had a little too much pepper on their shots, but it was fixed pretty quickly.

The speed trap itself is a little clumsy to use, mainly due to the length (a smidge under a meter) between the two photogates. I think we had to test 233 about fifteen times because we couldn't get a reading without balls whacking the exit of the chute. :ahh:

[Ellery]

At FLR they pretty much tested everyone it seemed, I know we were tested twice already at 11.27m/s so it would have been nice if they made the plans available for teams to build the same measuring device that they are being checked with earlier so to avoid the unncecesary guesses and tension during practice / inspection day.

Just my 2 cents to make things more efficient.

Ellery

[Scott358]

Based upon the fact that it's very difficult to measure the speed for most teams (we had theoretical values, tested per the angle/distance method, and even purchased a radar gun.. but even then tested (at our own request) a bit too high), why would the muzzle velocity not be a part of the robot/safety inspection (obviously only for those with a shooter).

With the exception of those who would change the value after it's checked, this would ensure compliance for those that wish to play by the rules. As it is now, many teams who probably think they are in compliance may not be, and hopefully they won't have issues after the fact.

I would suggest this be an added inspection check at Nationals.

[ChrisH]

Based upon the fact that it's very difficult to measure the speed for most teams (we had theoretical values, tested per the angle/distance method, and even purchased a radar gun.. but even then tested (at our own request) a bit too high), why would the muzzle velocity not be a part of the robot/safety inspection (obviously only for those with a shooter).

With the exception of those who would change the value after it's checked, this would ensure compliance for those that wish to play by the rules. As it is now, many teams who probably think they are in compliance may not be, and hopefully they won't have issues after the fact.

I would suggest this be an added inspection check at Nationals.

You might notice that all the posts before you were from about two weeks ago. At the LA regional last weekend every shooter went under the speed gun. If you couldn't pass or if you ripped up the inspection ball, you were not allowed to shoot. They will prpobably do the same at all the other regionals next week end and again in Atlanta.

[Al Skierkiewicz]

After a day of practice, the referees become very good at determining ball speed. Those that are suspect get tested out on practice day. Of all the robots I tested in two regionals, only one was fast, every other one was under and averaged about 11M/s.

[Scott358]

At the LA regional last weekend every shooter went under the speed gun. If you couldn't pass or if you ripped up the inspection ball, you were not allowed to shoot. They will prpobably do the same at all the other regionals next week end and again in Atlanta.

It's good that at some regionals this is happening, but as far I as know, as of now, it's up to the regional how they handle it. My team has been to 2 regionals and there was minimal inspection on this, but shooters that were definitely questionable.

As you said, perhaps everyone will be checked at Nationals, but until FIRST makes it part of the official inspection, it's only by luck that there will be consistent enforcement. Based upon the fact that this is factor that can give a significant advantage to those who operate above the speed and don't ever get checked, or can be an issue to those who are found to be above the speed after a match, I would still suggest it should be part of the official inspection specified by FIRST (and not left up to chance).

[hillale]

The 1 meter long speed tester at the St. Louis Regional was a joke. After breaking their tester by not lining it up correctly and hitting it with a ball, we got inconsistent readings all day. After finding a speed that the sensor said we could use, we tested our distance and we couldn't shoot 10 feet. If they are going to require speed testing at nationals, I should hope that they will come up with a more acurate way of testing.

[Al Skierkiewicz]

We checked the official FIRST speedo against two team's devices and all agreed. If you have used a radar gun to calibrate then your speed is way off. We have not found a radar gun that accurately read the ball speed from any direction. There is not enough surface to reflect a readable signal back to the gun to be accurate.

[Richard Wallace]

The 1 meter long speed tester at the St. Louis Regional was a joke. After breaking their tester by not lining it up correctly and hitting it with a ball, we got inconsistent readings all day. After finding a speed that the sensor said we could use, we tested our distance and we couldn't shoot 10 feet. If they are going to require speed testing at nationals, I should hope that they will come up with a more acurate way of testing.I don't have nearly enough data to reach a conclusion on this, but the following experiences might be helpful to those who perform ball speed tests, or to teams with robots that require ball speed testing, at the remaining events.

At STL, the head ref (Frank Kusiak, who was also head ref at MWR the following week) and the lead robot inspector (I) met Thursday to develop a procedure by which ball speed tests would be conducted. It went like this:

1. No ball speed testing was included in the required robot inspection checklist. This is established by FIRST headquarters and is not a regional option for lead robot inspectors.

2. Our interpretation of <S02> was that the head ref is the only official at the event who can order a ball speed test. The head ref may rely on his/her own observations and those of the other referees when deciding to order a test, or he/she may choose to respond to protests from competitors; i.e., drivers and/or HPs on the field.

3. When a ball speed test is ordered, the test equipment provided by FIRST is to be used in a manner directed by the FTA at the event. Robot inspectors will conduct the test with the assistance of the FTA if required, and with the cooperation of the team whose ball speed is being tested. This ensures that the team receives immediate feedback from the test.

4. Test conditions are as follows: the team is instructed to fire one ball per test through the FIRST ball speedometer into a curtain, and the speed of that ball is read using LabView. At STL we had the team repeat this test with several (at least five) balls, and looked for a consistent result. We ignored measurements that were inconsistent, assuming that high results were due to the shooter contacting the ball at the hard spot and that low results were due to incorrect speedometer alignment with the ball launch path.

5. When a consistent result exceeding 12 m/s was obtained, we advised the team that they needed to take corrective action. Teams were permitted to make software corrections at the test area and repeat testing immediately.

6. When a consistent result at 12 m/s or lower was obtained we advised the team and the head ref that the ball speed test had been passed.

Six teams had ball speed tests ordered by the head ref at STL. The only team for which multiple tests were ordered was 1625. My opinion is that 1625's shooter was legal in all matches that I observed; however, the backspin imparted by their shooter mechanism created a relatively flat trajectory and increased range, making their initial ball speed appear faster than it actually was. It was this appearance (IMO) that caused the head ref to order repeated ball speed testing. Again IMO, time spent on testing and overcorrection of the shooter wheel PWM level caused 1625 to perform worse than they could have in a few qualifying matches.

Two weeks later while volunteering as a robot inspector at Waterloo, I saw team 1596 suffer similar results for similar reasons: backspin caused flat trajectory and overcorrection caused low initial ball speed (with significant loss of range) in one or two qualifying matches. As with 1625, my opinion is that 1596's shooter was legal in all matches that I observed. In each case (1625 and 1596) the robot had one of the best (arguably the best) shooter at the event.

The point here is that it does not matter what the inspector or lead inspector thinks, only what he/she measures. What matters is the head referee's judgment, and in my opinion that judgment was remarkably consistent at events I attended and at those I watched via webcast.

My feeling is that there must be some method for ensuring compliance with <S02> and that the head referee's judgment is the best method. Test equipment and procedures can certainly be improved, and hopefully FIRST will provide some guidance.

[hillale]

I don't have nearly enough data to reach a conclusion on this, but the following experiences might be helpful to those who perform ball speed tests, or to teams with robots that require ball speed testing, at the remaining events.

At STL, the head ref (Frank Kusiak, who was also head ref at MWR the following week) and the lead robot inspector (I) met Thursday to develop a procedure by which ball speed tests would be conducted. It went like this:

1. No ball speed testing was included in the required robot inspection checklist. This is established by FIRST headquarters and is not a regional option for lead robot inspectors.

2. Our interpretation of <S02> was that the head ref is the only official at the event who can order a ball speed test. The head ref may rely on his/her own observations and those of the other referees when deciding to order a test, or he/she may choose to respond to protests from competitors; i.e., drivers and/or HPs on the field.

3. When a ball speed test is ordered, the test equipment provided by FIRST is to be used in a manner directed by the FTA at the event. Robot inspectors will conduct the test with the assistance of the FTA if required, and with the cooperation of the team whose ball speed is being tested. This ensures that the team receives immediate feedback from the test.

4. Test conditions are as follows: the team is instructed to fire one ball per test through the FIRST ball speedometer into a curtain, and the speed of that ball is read using LabView. At STL we had the team repeat this test with several (at least five) balls, and looked for a consistent result. We ignored measurements that were inconsistent, assuming that high results were due to the shooter contacting the ball at the hard spot and that low results were due to incorrect speedometer alignment with the ball launch path.

5. When a consistent result exceeding 12 m/s was obtained, we advised the team that they needed to take corrective action. Teams were permitted to make software corrections at the test area and repeat testing immediately.

6. When a consistent result at 12 m/s or lower was obtained we advised the team and the head ref that the ball speed test had been passed.

Six teams had ball speed tests ordered by the head ref at STL. The only team for which multiple tests were ordered was 1625. My opinion is that 1625's shooter was legal in all matches that I observed; however, the backspin imparted by their shooter mechanism created a relatively flat trajectory and increased range, making their initial ball speed appear faster than it actually was. It was this appearance (IMO) that caused the head ref to order repeated ball speed testing. Again IMO, time spent on testing and overcorrection of the shooter wheel PWM level caused 1625 to perform worse than they could have in a few qualifying matches.

Two weeks later while volunteering as a robot inspector at Waterloo, I saw team 1596 suffer similar results for similar reasons: backspin caused flat trajectory and overcorrection caused low initial ball speed (with significant loss of range) in one or two qualifying matches. As with 1625, my opinion is that 1596's shooter was legal in all matches that I observed. In each case (1625 and 1596) the robot had one of the best (arguably the best) shooter at the event.

The point here is that it does not matter what the inspector or lead inspector thinks, only what he/she measures. What matters is the head referee's judgement, and in my opinion that factor has been remarkably consistent at events I attended and at those I watched via webcast.

My feeling is that there must be some method for ensuring compliance with <S02> and that the head referee's judgement is the best method. Test equipment and procedures can certainly be improved, and hopefully FIRST will provide some guidance.

Thank you very much for this quote. You are correct in everything that you said. Very gracious and very professional. I should have taken more time to think before I posted.

[ChrisH]

Assuming the measured length is 1 meter, there is a simple calibration that can be performed to check the speed tester. Simply hold it so the ball can drop through it vertically. Hold the ball close to the first sensor and drop it through. The ball speed should read 2.2 m/s.

Don't belive me? do the math! Hint d=(at^2)/2. If anybody needs more than that post here and I'll walk you through it. I figured this out for our own tester.

ChrisH

[Richard Wallace]

Assuming the measured length is 1 meter, there is a simple calibration that can be performed to check the speed tester. Simply hold it so the ball can drop through it vertically. Hold the ball close to the first sensor and drop it through. The ball speed should read 2.2 m/s.

Don't belive me? do the math! Hint d=(at^2)/2. If anybody needs more than that post here and I'll walk you through it. I figured this out for our own tester.

ChrisHThe ball speedometer we used at STL and Waterloo had ~34.5 inch (0.876 meter) sensor spacing. For that spacing the correct speed reading for a ball released vertically just above the top sensor, with no initial velocity, would be 0.876 / sqrt( 2*(0.876)/9.8) = 2.07 meters per second. [edited several hours later; denominator is still the elapsed time, new numerator is the drop distance (incorrectly used 1 meter in original post)]

Be sure to measure the sensor spacing, and check that the actual spacing is entered in appropriate field of the LabView screen that controls the ball speedometer.

[TubaMorg]

Uh without actually checking anyones math, how is it that the speed of a 1 m drop is 2.2 m/s and the speed of a .876 m drop is 2.37 m/s ?

[Richard Wallace]

Uh without actually checking anyones math, how is it that the speed of a 1 m drop is 2.2 m/s and the speed of a .876 m drop is 2.37 m/s ?Sorry, my bad. Posted in haste earlier, then went elsewhere for several hours. Corrected now, I think.