Continuous Floor Loading?

[Orion.DeYoe]

A couple of things relating to floor loaders that I would like to get some options and alternate views on.
For my own use I like to categorize floor loading into two groups: "Continuous" and "Grabber" mechanisms.
Essentially a continuous floor loader is a system of belts or rollers that spin and feed a gamepiece continuously from the floor to wherever the gamepiece is stored in the robot. It doesn't require any positional actuation of manipulators while the gamepice is in transit to work.
A grabber floor loader is something that doesn't have any continuous components and requires properly timed actuation of the mechanism in order to pick up a gamepiece. These could be passive (like the scoop pickups we saw in 2013) or could be a gripper-type mechanism.
Firstly, are these two groups similar to how anyone else would classify floor loading mechanisms?
My main inquiry is about a third group. This group is a hybrid of continuous and grabber mechanisms. In these mechanisms a roller/belt pulls one or more gamepieces into a temporary holding position while the mechanism is actuated into a position where the roller (or some other mechanism) can release the gamepiece(s) into the real holding/staging area. A prime example of this would be 67's floor loader from 2012. I would place 118's 2011 mechanism in this class. My favorite from this year is 2614's pickup.
I have my own criteria for when to use each type however I want to know when have/would you use each type of floor loader? What are the advantages of each? What are some high-quality examples (if you could link to images/video that would be great)?
In addition, the hybrid style seems to be underestimated and seems to balance function with complexity very well. As I have never designed, built, or worked with one, what are certain situations where this style should NOT be used?

[Joe G.]

What you are calling "continuous" systems have been the best choice for the majority of FIRST games, because they by nature can be faster and reduce the need for precision and driver skill. If you can build a quality one, it will generally perform better in the long run.

Hybrid systems like 118 2011 are designed to overcome a discrepancy in the conditions between how an object must be gathered versus how it is released. In many cases, this difference can be lack of precision desired as a quality for pickup, with high precision being required for release in a pick and place game. Other examples include carefully controlled feeding (most games that involve shooters), or speed (movable jaw roller claws in 2011 for fast release). When such conditions are not in place, it's just needless complexity.

Non-continuous intake solutions have limited utility. Looking back on FIRST history, the only game where I would consider a passive or actuated intake to be indisputably optimal over a continuous one would be 1992, where the sheer number and density of balls made scoop pickups an effective solution. They've been made to work by many great teams in several games since, but very few designs beat the driver friendly nature of a roller based intake. They're often a much better choice for human loaded elements, but you asked about floor loading specifically.

[Orion.DeYoe]

Originally Posted by Joe G.

What you are calling "continuous" systems have been the best choice for the majority of FIRST games, because they by nature can be faster and reduce the need for precision and driver skill. If you can build a quality one, it will generally perform better in the long run.

Hybrid systems like 118 2011 are designed to overcome a discrepancy in the conditions between how an object must be gathered versus how it is released. In many cases, this difference can be lack of precision desired as a quality for pickup, with high precision being required for release in a pick and place game. Other examples include carefully controlled feeding (most games that involve shooters), or speed (movable jaw roller claws in 2011 for fast release).

Non-continuous intake solutions have limited utility. Looking back on FIRST history, the only game where I would consider a passive or actuated intake to be indisputably optimal over a continuous one would be 1992, where the sheer number and density of balls made scoop pickups an effective solution. They've been made to work by many great teams in several games since, but very few designs beat the driver friendly nature of a roller based intake. They're often a much better choice for human loaded elements, but you asked about floor loading specifically.

I agree. I always push for continuous mechanisms in design discussions.
Just to clarify, I consider roller grippers from 2011 continuous mechanisms, not hybrid, as the gripper is the final storage mechanism for the gamepiece.
The thing is that hybrid mechanisms solve many geometrical problems that make designing continuous floor loaders, storage elevators, and feeding systems difficult at times. Primarily it allows you to lift the gamepiece over the bumper without a complex system of belts. 67's 2012 floor loader was a roller and a bar that the balls were squeezed against.
The only time I can see this not being superior to a continuous mechanism is in cases where there are a large number of gamepieces. To the point where your mechanism can't hold a full load of gamepieces or can't transfer the contents of the mechanism before it's time to pick up more.

[randantor]

I would call our 2014 bot's intake a good example of a 'hybrid' pickup. We use a roller to grip the ball, but once the ball goes through the roller the roller needs to go back up to load the ball into the shooter. We can also hold one ball inside the roller while another is loaded into the shooter, which we used during autonomous. There's a good video of the pickup sequence on our prototype robot here.

In our case it was successful in spite of being fairly slow because there was only ever one game piece to pick up. Once our bot got the ball in the roller and off the ground, we could be driving down the field while it loaded the rest of the way. If the game were more similar to 2013 or 2012 where there were many game pieces in play at the same time, it likely wouldn't have worked out as well.

The pickup sequence is also automated, so the drivers do not need to be concerned with the multiple actuations needed to pick up a ball. I would say that that automation is essential to any complicated form of pickup mechanism to ease the load on the driver.

[DampRobot]

The last time I can think of a "non-continuous" mechanism being the most effective solution was 2005. Since then, as others have mentioned, roller systems have been most effective, as they make it much easier to pick up an object because they reduce the required accuracy.

Interestingly, end game manipulators tend to be what you'd call "non-continuous."

[EricH]

It depends on the application, of course. However, in general, there are also two types of applications.

1) Precision Placement. 2005, 2007, 2011 come to mind, as well as the 2008 endgame. In this, you're trying to place something on something else with a reasonable degree of accuracy.

2) In the Big Hole. Shooting-type games tend to land in this category, just put a lot of projectiles on the target.


Generally, precision placement games will lend themselves for a gripper-type pickup. (If you're going to lump roller claws into "continuous", you need to make sure that you're ONLY lumping the rollers in to use your own definition consistently. I'd consider them hybrid; it's a gripper with a continuous in-feed.) And yes, some of the best pick-and-place robots out there have used gripper-type claws without any sort of continuous component.

On the other hand, "big hole" games will tend to call for a continuous pickup to acquire ammunition for the massive shot at the target. Aerial Assist, due to only one gamepiece, was a slight exception, but most of the successful teams tended to use continuous.

And then there are the hybrids. Often, a given challenge will call for something that neither grippers nor continuous can completely provide. That's when life gets interesting, as you try to figure out whether to add a continuous to a gripper or vice versa, and automate the whole thing... I'll pick on 1197 this last year as having a hybrid. The initial intake was continuous, but if the storage/secondary intake wasn't in the right place the ball was then useless, putting it more in a gripper class. (Some clever code made the positioning automatic, or very nearly automatic, when the driver pressed the right button.) If both were in the right place, everybody within 15 feet of the sides of the field knew when they hit a ball they wanted as it smacked into the back of the storage...

[SoftwareBug2.0]

One other way that you could classify pickup systems is by whether they handle more than one of the item in the same part at the same time or if they must be dealt with in a sequence. This is orthagonol to whether they're dealt with continously or in a batch.

I agree the previous posters that continous is usually better. I am especially inclined to agree with Joe G.'s reasoning that they're easier to control. Often it's trivial: just turn it on and let it run.

[glennword]

Don't forget about 1986's 2013 pickup, with the "thumbs" for indexing and the arm. I would classify that as hybrid, although the argument could be made that it was purely passive or noncontinuous.
https://www.youtube.com/watch?v=uKy-IKDq_6o

[Joseph Smith]

In my opinion, the choice between a continuous pickup and a singular pickup comes down to one factor- how many game pieces your robot can possess.
If you can ever possess more than one game piece, I would always go for the fastest pickup possible, which is always going to be a continuous pickup. If you are limited to possessing one piece at a time, then you can consider a singular pickup, but this isn't necessarily always the best option. When it comes to acquiring and possessing gamepieces, the fastest and most reliable loader is the best loader.

[Orion.DeYoe]

Originally Posted by glennword

Don't forget about 1986's 2013 pickup, with the "thumbs" for indexing and the arm. I would classify that as hybrid, although the argument could be made that it was purely passive or noncontinuous.
https://www.youtube.com/watch?v=uKy-IKDq_6o

That was a completely passive mechanism. No active components (other than the drivetrain) were involved in getting the gamepiece into the scoop.

[BBray_T1296]

Originally Posted by Orion.DeYoe

That was a completely passive mechanism. No active components (other than the drivetrain) were involved in getting the gamepiece into the scoop.

How about the mechanism for lifting the scoop from the floor to the hopper?