A better way of determining if the puck crossed the goal line

[IPostedWhat]

presumably these new chip doodads (to use the technical term) would be used as a back-up, like video review is now. Primary indication of a goal or not is the referee (and then the useless guy taking up a $200 first row seat).

Change doodads to doohickeys, and I'm digging this post.

 

[Slimpikins]

What if we made the puck glow and then if it crossed the line a bunch of spot lights would shine down on the net or maybe some flames would shoot off from the posts!

Ya, that would rock. Get Fox on the phone!

Hahaha, that will definitely get some NASCAR fans over to the NHL.

 

[KennyBanyeah]

I don't know that it would be necessary to hire an engineer to recalibrate it every period. The programming's what requires an engineer, but calibrating instruments that have already been programmed can be pretty simple.

I also think the technology has advanced enough since 1998 that sensors along the goalposts and crossbar would be unnecessary. Vertically oriented sensors in the ice along the goal line should suffice, so the sensor wouldn't have to deactivate every time the net was off its moorings. Now we're just talking about adding a third dimension to a well-known model of an object with radius r intersecting a line. That's very doable.

At this point, the only things holding up this project are the durability of the in-puck electronics, the expense of the sensors, and whether or not the human element should remain in the game.

Personally, I think that getting it right is more important than the human element, and the human element will still exist in cases like judgement calls on penalties and other things.

All good points.

However the way you pointed out the radius in scientific terms made me realize another problem.

There could be no constant r in an equation like this as we are not dealing with a round object.

The distance from the centre of the puck would vary with the obliquity of the puck as it crosses the line / sensor(s). When on edge the "r" would be somethin like 0.75 cm. When lying flat it would be closer 2 cm. Again, the issue of resolution would come up. We're dealing with a matter of millimetres here. I still think that it would be expensive to make a system that has a greater resolution than the human eye.

This discussion has led me to another problem. If you only had sensors in the ice, how many sensors would you need across the goal line to achieve the desired level of accuracy?

IMO, the three dimensional nature of hockey, the "un-roundness" of the puck and the scale with which we are working makes this a problem that is not practical to the structure and budget of 30 NHL rinks.

 

[Darkstone42]

All good points.

However the way you pointed out the radius in scientific terms made me realize another problem.

There could be no constant r in an equation like this as we are not dealing with a round object.

The distance from the centre of the puck would vary with the obliquity of the puck as it crosses the line / sensor(s). When on edge the "r" would be somethin like 0.75 cm. When lying flat it would be closer 2 cm. Again, the issue of resolution would come up. We're dealing with a matter of millimetres here. I still think that it would be expensive to make a system that has a greater resolution than the human eye.

This discussion has led me to another problem. If you only had sensors in the ice, how many sensors would you need across the goal line to achieve the desired level of accuracy?

IMO, the three dimensional nature of hockey, the "un-roundness" of the puck and the scale with which we are working makes this a problem that is not practical to the structure and budget of 30 NHL rinks.

For the shape of the puck, just use a cylindrical graphing system when you program it. They do it in electro-physics all the time to deal with wires. The degree of symmetry is still high, and the sensor can detect angle. It's a little more complicated than the problem for a sphere, but not that much.

As for the sensors, you could just scale up the design of a flat laser. It would only be one sensor. Most people deal with lasers that are just a dot, but if you use a low-scattering laser with a "flat" beam, you could accomplish the same thing. Like the ones they use at Wal Mart and Lowe's and stuff.

Or you could design it so that there are a few sensors located in various locations around the rink, which provide a constant computer image of the rink, and the sensor in the puck could just tell you where the puck is relative to the other sensors. Honestly, I think this is the way to go. Stick two to three sensors in the goal, one just inside each post and the other in the middle, and they will provide you with an image of the goal line. Then a computer can track the puck relative to the goal line.

 

[sportfan09]

think someone could come up with wii p@rn,being married has it's moments

 

[Eddie_Shack]

I can already see the goalies lining their gloves and pads with aluminum foil.

 

[KennyBanyeah]

For the shape of the puck, just use a cylindrical graphing system when you program it. They do it in electro-physics all the time to deal with wires. The degree of symmetry is still high, and the sensor can detect angle. It's a little more complicated than the problem for a sphere, but not that much.

As for the sensors, you could just scale up the design of a flat laser. It would only be one sensor. Most people deal with lasers that are just a dot, but if you use a low-scattering laser with a "flat" beam, you could accomplish the same thing. Like the ones they use at Wal Mart and Lowe's and stuff.

Or you could design it so that there are a few sensors located in various locations around the rink, which provide a constant computer image of the rink, and the sensor in the puck could just tell you where the puck is relative to the other sensors. Honestly, I think this is the way to go. Stick two to three sensors in the goal, one just inside each post and the other in the middle, and they will provide you with an image of the goal line. Then a computer can track the puck relative to the goal line.

How would a cylindrical model account for a disc shaped puck? Without doing a three dimensional model you have no chance. This eliminates the "only having sensors in the ice" model.

If, as you say, they have a sensor that deals with angle (or obliquity), then I would be VERY curious as to how quickly the puck-cored sensor can deal with that change in angle/aspect/obliquity (now we are dealing with a 3-d , and VERY expensive sensor; we're talking beyond anything NASA has done). I'd love to see that happen. This will bring a temporal aspect into the movement of the puck.

As for having sensors around the rink, we again, return to the problem of having the hockey environment being unsuitable (because of the ever-changing nature of modern arenas) to this kind of data gathering.

This kind of decision making will never surpass what is currently in place, IMO.

 

[Darkstone42]

How would a cylindrical model account for a disc shaped puck? Without doing a three dimensional model you have no chance. This eliminates the "only having sensors in the ice" model.

If, as you say, they have a sensor that deals with angle (or obliquity), then I would be VERY curious as to how quickly the puck-cored sensor can deal with that change in angle/aspect/obliquity (now we are dealing with a 3-d , and VERY expensive sensor; we're talking beyond anything NASA has done). I'd love to see that happen. This will bring a temporal aspect into the movement of the puck.

As for having sensors around the rink, we again, return to the problem of having the hockey environment being unsuitable (because of the ever-changing nature of modern arenas) to this kind of data gathering.

This kind of decision making will never surpass what is currently in place, IMO.

The cylindrical graphing system is a three-dimensional graphing system. It's three components are the radius, s, the length, z, and the angle, theta. And the sensor need only indicate rotational angle and position, which could be communicated separately. It would send one message saying that the midpoint of the puck is at such and such location, and another piece of information giving the orientation. Then the computer can calculate whether or not the puck is across the goal line.

I think the only technology that doesn't exist is the technology of implementation. Specifically, the soft-material computer chip.

 

[elocomotive]

If, as you say, they have a sensor that deals with angle (or obliquity), then I would be VERY curious as to how quickly the puck-cored sensor can deal with that change in angle/aspect/obliquity (now we are dealing with a 3-d , and VERY expensive sensor; we're talking beyond anything NASA has done). I'd love to see that happen. This will bring a temporal aspect into the movement of the puck.

It's not beyond anything NASA has done... it's not even beyond anything Nintendo has done. And it can't be THAT expensive b/c those controllers don't cost that much and I'm sure they have a HUGE markup.

The sensor doesn't have to do anything except relay where it is and the angle. The computer would create the 3D model of the goal area and show whether the puck has gone across or not.

You guys seem to hung up on some kind of sensors the puck sets off - like a trip wire. It would be much easier to simply calibrate the puck to a point on the ice (like a GPS device, but in a rink-sized world) and measure all data based on it's distance from that point. A computer could be incredibly accurate. The Hawkeye system in tennis uses cameras to relay data (not even a chip involved) - something like that is certainly possible with or without a chip. Keep in mind this system is also used in an environment where millimeters are important and speeds beyond (120-140 mph+) the speed a puck moves.

And while this would cost some R&D, none of the technologies involved are not in commercial use already - GPS devices, video game controllers, 3D imaging, etc. It's just combining them in a way that works for hockey.

 

[Darkstone42]

It's not beyond anything NASA has done... it's not even beyond anything Nintendo has done. And it can't be THAT expensive b/c those controllers don't cost that much and I'm sure they have a HUGE markup.

The sensor doesn't have to do anything except relay where it is and the angle. The computer would create the 3D model of the goal area and show whether the puck has gone across or not.

You guys seem to hung up on some kind of sensors the puck sets off - like a trip wire. It would be much easier to simply calibrate the puck to a point on the ice (like a GPS device, but in a rink-sized world) and measure all data based on it's distance from that point. A computer could be incredibly accurate. The Hawkeye system in tennis uses cameras to relay data (not even a chip involved) - something like that is certainly possible with or without a chip. Keep in mind this system is also used in an environment where millimeters are important and speeds beyond (120-140 mph+) the speed a puck moves.

And while this would cost some R&D, none of the technologies involved are not in commercial use already - GPS devices, video game controllers, 3D imaging, etc. It's just combining them in a way that works for hockey.

This is kind of what I'm looking at.

Of course, for hockey, I want the review display to include the coding and algorithms through which the computer is going. But that's because I like that kind of stuff.

 

[Ho_Brah]

didn't think this thread would get this much input, glad it did, some very interesting discussions!

 

[Slimpikins]

Here is why this will never happen, cost benefit ratio. Look at the cost it would take to develop, produce, and implement all of this and what is the benefit? 30 goals a year are overturned? 50 goals? 100 goals a year are overturned at a cost of hundreds of millions of dollars from a league that is already struggling to turn a profit from half the league?

No way it is even on the radar. It is a fun discussion but the marginal benefit is not worth the steep price.

 

[KennyBanyeah]

It's not beyond anything NASA has done... it's not even beyond anything Nintendo has done. And it can't be THAT expensive b/c those controllers don't cost that much and I'm sure they have a HUGE markup.

The sensor doesn't have to do anything except relay where it is and the angle. The computer would create the 3D model of the goal area and show whether the puck has gone across or not.

You guys seem to hung up on some kind of sensors the puck sets off - like a trip wire. It would be much easier to simply calibrate the puck to a point on the ice (like a GPS device, but in a rink-sized world) and measure all data based on it's distance from that point. A computer could be incredibly accurate. The Hawkeye system in tennis uses cameras to relay data (not even a chip involved) - something like that is certainly possible with or without a chip. Keep in mind this system is also used in an environment where millimeters are important and speeds beyond (120-140 mph+) the speed a puck moves.

And while this would cost some R&D, none of the technologies involved are not in commercial use already - GPS devices, video game controllers, 3D imaging, etc. It's just combining them in a way that works for hockey.

This is a more interesting idea. I like it.

I still don't see why it's necessary, but it makes sense if the NHL were to go that way.

 

[KennyBanyeah]

It's not beyond anything NASA has done... it's not even beyond anything Nintendo has done. And it can't be THAT expensive b/c those controllers don't cost that much and I'm sure they have a HUGE markup.

The sensor doesn't have to do anything except relay where it is and the angle. The computer would create the 3D model of the goal area and show whether the puck has gone across or not.

You guys seem to hung up on some kind of sensors the puck sets off - like a trip wire. It would be much easier to simply calibrate the puck to a point on the ice (like a GPS device, but in a rink-sized world) and measure all data based on it's distance from that point. A computer could be incredibly accurate. The Hawkeye system in tennis uses cameras to relay data (not even a chip involved) - something like that is certainly possible with or without a chip. Keep in mind this system is also used in an environment where millimeters are important and speeds beyond (120-140 mph+) the speed a puck moves.

And while this would cost some R&D, none of the technologies involved are not in commercial use already - GPS devices, video game controllers, 3D imaging, etc. It's just combining them in a way that works for hockey.

It sure is as regards resolution. Nintendo has never had resolution CLOSE to 1 mm.

 

[elocomotive]

This is kind of what I'm looking at.

Of course, for hockey, I want the review display to include the coding and algorithms through which the computer is going. But that's because I like that kind of stuff.

The really awesome thing for tennis is that it doesn't just show the ball hitting, but the ball actually stretching out and changing shape as it picks up speed and then flattening out and extending as it hits the ground. The ball is about 4 inches wide, but it contacts with the ground for 5-6 inches or so before the recoil pushes it back up again. It's cool stuff.

 

[elocomotive]

Here is why this will never happen, cost benefit ratio. Look at the cost it would take to develop, produce, and implement all of this and what is the benefit? 30 goals a year are overturned? 50 goals? 100 goals a year are overturned at a cost of hundreds of millions of dollars from a league that is already struggling to turn a profit from half the league?

No way it is even on the radar. It is a fun discussion but the marginal benefit is not worth the steep price.

I think you guys are overestimating the cost here. I don't know when it happens, but I guarantee this happens in every sport. It's just a matter of when. I would guess baseball (most conservative of all the sports) will probably be the last to institute it.

It doesn't even have to be as complicated as we're talking about either. It could just be the technology FOX used where it colors the puck to allow you to see it through objects (like a sprawled out goalie). That technology existed more than a decade ago.

Time has showed us a steady march of human jobs converting to computer/machine replacement. I'll bookmark this thread and in 2027 you'll all be eating some crow!

 

[Destroydacre]

It's not beyond anything NASA has done... it's not even beyond anything Nintendo has done. And it can't be THAT expensive b/c those controllers don't cost that much and I'm sure they have a HUGE markup.

The sensor doesn't have to do anything except relay where it is and the angle. The computer would create the 3D model of the goal area and show whether the puck has gone across or not.

You guys seem to hung up on some kind of sensors the puck sets off - like a trip wire. It would be much easier to simply calibrate the puck to a point on the ice (like a GPS device, but in a rink-sized world) and measure all data based on it's distance from that point. A computer could be incredibly accurate. The Hawkeye system in tennis uses cameras to relay data (not even a chip involved) - something like that is certainly possible with or without a chip. Keep in mind this system is also used in an environment where millimeters are important and speeds beyond (120-140 mph+) the speed a puck moves.

And while this would cost some R&D, none of the technologies involved are not in commercial use already - GPS devices, video game controllers, 3D imaging, etc. It's just combining them in a way that works for hockey.

The Hawk eye system doesn't work if the cameras can't see the ball. Many goal/no goal controversies come from not being able to see the puck. As for a gps type system, well those aren't 100% accurate either. I'd sure hate for my goalie to glove the puck on the goal line but for the gps system to say the puck was in.

 

[Slimpikins]

I think you guys are overestimating the cost here. I don't know when it happens, but I guarantee this happens in every sport. It's just a matter of when. I would guess baseball (most conservative of all the sports) will probably be the last to institute it.

It doesn't even have to be as complicated as we're talking about either. It could just be the technology FOX used where it colors the puck to allow you to see it through objects (like a sprawled out goalie). That technology existed more than a decade ago.

Time has showed us a steady march of human jobs converting to computer/machine replacement. I'll bookmark this thread and in 2027 you'll all be eating some crow!

The Foxtrax they used to highlight the puck with was nowhere near accurate enough to detect if a puck has completely crossed the line. To develop the software and hardware, produce it, and test it thoroughly enough to feel comfortable putting the outcome of a Stanley Cup on it's accuracy and effectiveness will absolutely cost multi millions of dollars. That doesn't include the price to install and maintain the equipment.

Annexing fourteen homeowners into an HOA after the previous builder forged some documents has cost over 100,000 in legal fees alone. With a cost like that to settle a relatively simple matter I am very confident that the cost of the proposed goal detection system will come with a hefty price tag.

It isn't as simple as pulling a chip out of a video game controller and gluing it to the back of a puck. Make no mistake it is a massive undertaking.


EDIT: I'm drunk, hopefully that all made sense...

 

[elocomotive]

The Hawk eye system doesn't work if the cameras can't see the ball. Many goal/no goal controversies come from not being able to see the puck. As for a gps type system, well those aren't 100% accurate either. I'd sure hate for my goalie to glove the puck on the goal line but for the gps system to say the puck was in.

Yes, I know. It was only an example of how you could display the visual based on computer data with a non-spherical object (as noted earlier).

And yes, Fox Trax was not 100% accurate. But Fox Trax was (a) used in real time, not on a playback where the computer would have some precious fractions of a second to computer the data and (b) more than a decade ago, surely there are constantly advances, etc. And I'd also note the human eye is far from 100% accurate, but I have little doubt they could devise a computer system that would be 99.9% accurate.

If you think the cost outweighs the need at the moment, well, that's a valid argument, though I'd note none of us know the amount of that cost and any potential savings in personnel it might create (if any). But it may be that cost (now, in the near future, for a long time) outweighs the need.

If you honestly believe that this is not technically possible, I'd say you either lack imagination, are ignorant of technology, or both. We fly aircraft from halfway around the world. I really don't think getting a chip in a puck so we can accurately know where it is is something we even need to argue about. Best way to do it - lots of options. Not accurate? Not possible? That's absurd.

 

[Destroydacre]

Yes, I know. It was only an example of how you could display the visual based on computer data with a non-spherical object (as noted earlier).

And yes, Fox Trax was not 100% accurate. But Fox Trax was (a) used in real time, not on a playback where the computer would have some precious fractions of a second to computer the data and (b) more than a decade ago, surely there are constantly advances, etc. And I'd also note the human eye is far from 100% accurate, but I have little doubt they could devise a computer system that would be 99.9% accurate.

If you think the cost outweighs the need at the moment, well, that's a valid argument, though I'd note none of us know the amount of that cost and any potential savings in personnel it might create (if any). But it may be that cost (now, in the near future, for a long time) outweighs the need.

If you honestly believe that this is not technically possible, I'd say you either lack imagination, are ignorant of technology, or both. We fly aircraft from halfway around the world. I really don't think getting a chip in a puck so we can accurately know where it is is something we even need to argue about. Best way to do it - lots of options. Not accurate? Not possible? That's absurd.

It's not that I don't think it's possible, I'm just not ready to trust technology as an absolute indication for determining goals, that's all. With as good as technology is now, there is need for improvement and we're both in agreement that it will cost money to develop. I just don't see the NHL springing the cost for that. Now if someone else can develop a system, then the NHL may very will pick it up and implement it. I think you believe this technology will be developed quicker than I believe it will. I think it's still a good ways away. But yeah, let's bring this back up in 2027 and see who was right! :p