Use of GPS for props

[Eleran]

I'd be waiting to see what happens when the first one decides to go rogue.

Edited March 26, 2016 by Eleran

[KVG_DC]

That's a VK revival show waiting to happen.

[drumcorpsfever]

That's a VK revival show waiting to happen.

Laughed till I cried! Stopped short of the goal line!

Edited for: It's first and goal from the 9-yard line!

Edited March 26, 2016 by drumcorpsfever

[DrumManTx]

I can't handle this. HAHAHAHA!

[KVG_DC]

I especially like the guy who rode the thing for 15 yards right where the stuff jamming the accelerator was and just bailed and let it go for someone else to handle

[Clutchtow]

I know... no motors in DCI... (note: I marched with a Corps that helped bring about the 'no animals" rule)... so with that aside, just as a proof-of-concept exercise, why not utilize a simple system similar to robotic studio TV cameras instead of GPS..?

Companies such as Vinten have had systems in play for decades that will move and relocated motorized camera supports based on a simple "direction and distance" model from a singular home starting location that may easily apply to platform and/or prop movement. Something similar would require nothing more than a simple definable 'target' for a starting reference. This would require no need for advanced venue scouting or individual re-calibration per new venue. You could utilize the fact that pretty much all performance areas (football fields) are standard and uniform in dimension to your advantage.

In theory, as the props or platforms were being moved towards the field in pre-show, you would define a repeatable and standard location as your initial targeting position. It could be something as simple as the very corner of the end zone markings where the sideline and end zone lines intersect. If you were to align a square stage with two sides lined up exactly along those two intersecting field markings, the prop would then have a base or 'home' reference that it could easily use to calculate and determine the location and dimensions of the performance area. It now has a defined X/Y axis area for direction. All it needs now is distance information for positioning. It can obtain that from something as simple as defined distance traveled based on calculated wheel travel. So... give it a home for reference... tell it what direction it needs to go based on that home reference... tell it how many wheel turns it needs to make to achieve the required distance... and BOOM... your stage is set at position #1 for the opening number. You then utilize each new position as the targeting location for the next move. It's kind of like reading directions to a scavenger hunt. "Start here... then go there... from there, go here..." and so on. Speed programming could be done by defining the 'time' used to complete the required distance traveled per move, easily incorporating prop movement into the visuals.

Again, just a thought... Something along those lines might provide a simpler engineering approach to a GPS based system. But hey... I'm no engineer, so I could be totally wrong on all this...

It would be certainly simpler, and this method is called dead reckoning. I used it a couple times for autonomous movement back in my FRC days. The issue is that there is minor slippage in the wheels and the encoders (the devices that count the wheel turns) that make it so each move would have some error to it, and by the time you get to the 5th set you could now be off by a yard. So of course you then try to correct it in programming, but once you get to a new field, maybe one with grass instead of turf, those corrections aren't valid. Also, say your batteries aren't charged fully, now the stages need more power to move at the same rate. That is correctable though via a PID control loop. But if you were on one field, and were willing to iron out all the external variables, you could definitely get this method to work!

Disclaimer: I am just a sophomore electrical engineering student with delusions of grandeur, so take this all with a grain of salt haha

[sadtrombonesound]

I think as long as you start them from the right position on the field (and that wouldn't be hard, just position them on the field like any other prop), you can pre-program the vectors for each one from a single computer, then you don't have to worry about keeping signals or using GPS or referencing points on the field or anything. Each device would just go through the motions (and hit its drill dots, to use a surprisingly accurate analogy) as it was programmed to do for 11 minutes.

I'm pretty sure that's what they did here with the quadcopters, but if it were just wheeled stuff on the field you would only need to worry about 2 dimensions, so it would be easier (maybe): https://www.youtube.com/watch?v=ShGl5rQK3ew

[Ghost]

I'm pretty sure that's what they did here with the quadcopters, but if it were just wheeled stuff on the field you would only need to worry about 2 dimensions, so it would be easier (maybe): https://www.youtube.com/watch?v=ShGl5rQK3ew

Might not be cost effective for a drill writer to use for ideas.

[deftguy]

I know... no motors in DCI... (note: I marched with a Corps that helped bring about the 'no animals" rule)... so with that aside, just as a proof-of-concept exercise, why not utilize a simple system similar to robotic studio TV cameras instead of GPS..?

Companies such as Vinten have had systems in play for decades that will move and relocated motorized camera supports based on a simple "direction and distance" model from a singular home starting location that may easily apply to platform and/or prop movement. Something similar would require nothing more than a simple definable 'target' for a starting reference. This would require no need for advanced venue scouting or individual re-calibration per new venue. You could utilize the fact that pretty much all performance areas (football fields) are standard and uniform in dimension to your advantage.

In theory, as the props or platforms were being moved towards the field in pre-show, you would define a repeatable and standard location as your initial targeting position. It could be something as simple as the very corner of the end zone markings where the sideline and end zone lines intersect. If you were to align a square stage with two sides lined up exactly along those two intersecting field markings, the prop would then have a base or 'home' reference that it could easily use to calculate and determine the location and dimensions of the performance area. It now has a defined X/Y axis area for direction. All it needs now is distance information for positioning. It can obtain that from something as simple as defined distance traveled based on calculated wheel travel. So... give it a home for reference... tell it what direction it needs to go based on that home reference... tell it how many wheel turns it needs to make to achieve the required distance... and BOOM... your stage is set at position #1 for the opening number. You then utilize each new position as the targeting location for the next move. It's kind of like reading directions to a scavenger hunt. "Start here... then go there... from there, go here..." and so on. Speed programming could be done by defining the 'time' used to complete the required distance traveled per move, easily incorporating prop movement into the visuals.

Again, just a thought... Something along those lines might provide a simpler engineering approach to a GPS based system. But hey... I'm no engineer, so I could be totally wrong on all this...

All of Vintens robo cams require a wiring harness, and cannot currently be used wirelessly. They are also notorious for not landing quite on target. I am not sure you can operate 5 of them off a single program unless you use a modified automation system like "Ignite" by Harris. When I worked at WABC, we could not move more than one robo cam at a time.

The wiring harnesses of robo cams are really sensitive - and cannot be stepped on, or even run over by the camera itself. I cannot see this working on the field unless a wireless system is developed.

Edited March 28, 2016 by deftguy

[skevinp]

but if it were just wheeled stuff on the field you would only need to worry about 2 dimensions

From a wheel's perspective I suspect the field is quite 3 dimensional indeed.