Shooter vs Dumper Trade Study

[Swampdude]

So what are the pros and cons of either design and why is one better than the other?

I'll start:
To me, shooting assumes you can't get to a trailer to score and you need to score from a distance (I guess because of the limited drive capabilities). But having 3 trailers to score in seems like there will always be one nearby. Also shooting seems like it will take too long and have more chances for jamming. Also in Aim High hitting a stationary target was hard enough, so how do you hit a smaller moving target with non-uniform objects (broken balls) consistently?

A dumper assumes you can get up close to a trailer, scores all at once and less mechanisms to jam. Doesn't care about ball shape.

I realize there's hybrids but ultimately it seems you still fall into one of these categories. Yes/No?

[Kellen Hill]

I personally think that effective dumpers will score a majority of the points and dominate the game. Unless you have a shooter like Wildstang in 2006 of course.

[MrForbes]

A short range turreted shooter will allow you to score when you are close to the goal, and from either side of your robot. A dumper probably requires you to be lined up and right next to the goal. Since the goals will probably be driving away from you, using the finest evasive maneuvering techniques, you'd have to be really lucky to be able to dump a load of balls in a functioning robot's trailer.

[ggoldman]

There are times during game play where a dump is very viable..and easy.

When another team is unloading into the airlock, the trailer is stationary.

When another team is at the outpost getting the empty cell, the trailer is stationary.

When another team is pushed up against a wall by two robots, the trailer is stationary.

A dumping bot of sorts only needs to get to a trailer goal once or twice in a match to make a big difference.

[Pjohn1959]

[quote=Swampdude;800801]So what are the pros and cons of either design and why is one better than the other?QUOTE]

So why not build something that does both effectively?

[Jonathan Norris]

How about a shooting dumper??

[Travis Hoffman]

Originally Posted by Jonathan Norris

How about a shooting dumper??

How about a turreted shooting dumper?

[Brandon Holley]

Originally Posted by Travis Hoffman

How about a turreted shooting dumper?

Bingo.

[Bongle]

Advantage to a shooter is that presumably, it can act as a dumper simply by slowing the spinning of its power mechanism (wheels, etc).

Based on our tests so far, it appears that tracking and auto-targeting opponent's trailers' is possible:
-We successfully wrote code that can distinguish an opponent's trailer
-The same code can give you distance and heading information within a few % points of the true value.
-By tracking the change in distance and heading, you can determine both the position and velocity of an opposing team's trailer, and thus lead it by an appropriate amount.

But:
-The enemy trailer will pretty much never be moving at a constant velocity
-There appears to be fairly substantial slowdowns in the camera code. It's not really real-time, more like 500ms-behind-real-time. We don't know how much this can be fixed, that'll be tonight's meeting's topic.
-The balls are irregular, and will become even more irregular as regionals proceed. Thus, muzzle velocities will be unpredictable.
-'Camping' and shooting will not be effective: you'll be a sitting duck for dumpers. Handling the accelerations that your robot is undergoing as well as the accelerations of your target will make your shooting much less accurate

When we were doing our original analysis, we essentially decided that shooters will have more opportunities to score, but will convert fewer of those opportunities into actual points.

[Donut]

There is a hybrid option that really doesn't fall into these two categories though. A shooter doesn't have to use a turret and camera (and may not want to since it restricts you to one ball at a time), but could instead pursue a similar strategy to a dumper of getting close to an opponent's goal and unloading a large amount of balls into it. Being a shooter allows you to power the balls (instead of relying on a gravity dump) and gives you a few foot range rather than requiring your robot to be right next to the opponent.

[Goldeye]

Originally Posted by Bongle

But:
-The enemy trailer will pretty much never be moving at a constant velocity
-'Camping' and shooting will not be effective: you'll be a sitting duck for dumpers. Handling the accelerations that your robot is undergoing as well as the accelerations of your target will make your shooting much less accurate

The low friction and thus low acceleration minimizes the effect of these problems.

[Bongle]

Originally Posted by Goldeye

The low friction and thus low acceleration minimizes the effect of these problems.

If your shooter has a moderate speed (8ft/s), then a robot just 10ft away could accelerate hard enough to be 1.5ft away from where you expected it to be (if you assume constant velocity). If you add in the inevitable imprecision of an orbit-ball shooter, the amount of balls you'd get in would be low.

A powered dumper seems to be the way to go.

[Goldeye]

Originally Posted by Bongle

If your shooter has a moderate speed (8ft/s), then a robot just 10ft away could accelerate hard enough to be 1.5ft away from where you expected it to be (if you assume constant velocity). If you add in the inevitable imprecision of an orbit-ball shooter, the amount of balls you'd get in would be low.

A powered dumper seems to be the way to go.

By similar arguments, a moving robot further than 1-2 feet away will, at the least, be a challenge to reach and dump into, particularly for long enough to dump all the stored orbit balls. I don't know the radius of the goal offhand. Isn't the radius about 1.5ft? If the error due to acceleration is just that, then the only real concern is the precision of a shooter. I can't comment on that

Also, 10 feet really isn't that close. Robots can easily get within 5 feet of eachother, I just envision trouble getting closer (and staying there)

There's no right answer, and teams will get both ways working, but I feel confident saying this just from the nature of this year's driving: if you can build a device that aims independently of your robot and propels the balls (at least slightly), you will have a much easier time getting balls in the goal. There are still other tradeoffs, of course.

[David Guzman]

In my opinion a shooter robot will be more reliable than a dumper robot.

While a dumper robot can score massive points when given the opportunity, it is relying on the mistakes from the opposing alliance to be able to make a dump.

A robot with a shooter will need to use the camera to track the trailers to be a threat. If the camera is used successfully and the shooter is geared correctly so it can score a lot of balls very quickly, then it will dominate.

I think the combination of a good shooter and a good dumper in an alliance will be very powerful.

[Mr B]

It seems to me that the biggest constraint of dumping balls into a goal is the size limit of the robot <R11>. Without being able to exceed the perimeter, dumpers would have to be right next to a goal and be relatively unmolested to score a high percentage of balls. Still, a team that figures out a way to dump a load reliably and fast would be a force.