# 2018 - How did teams choose the angle for their intake 'funnel'?

I have been doing a personal CAD exercise to design a 2018 robot. While looking at successful intake designs, I noticed teams like 254 and 1323 have a ‘funnel’ feature to their intake, allowing for a more effective touch-own of Power Cubes.

• What angle (shown in the image below) did you go with?
• What prototyping process did you use to decide on this angle?

4 Likes

For us this angle was the subject of a lot of prototyping and experimentation.
However, one thing to note is that 254/1323’s intakes also actuate open so the angle here is a bit less critical than ours for instance that was always closed.

5 Likes

The three cases you need to design for is.

1. (minimum) approaching a cube normal to a face.
2. (maximum) approaching a cube with an edge closest to the robot.
3. Approaching a cube that’s against the switch wall with the robot bumper also against the switch wall. (18 of the 22 cubes you could grab started in this position)

5 Likes

My memory is a bit fuzzy, but I recall approaching intake design as parameterization. Rather than choosing a specific angle that worked, we wanted to determine at what angle did a particular design NOT work. By identifying the extremes we then translated that into design constraints that could be integrated with packaging considerations. The optimal angle was then found via iteration over the course of the season (there’s a picture somewhere of the 8 or so sets of intake plates we went through).

It also helped that we identified pretty early on (the day after kickoff I think?) that we wanted to intake cubes at any angle in any orientation, and be able to either eject them in a direction or drop them from above.

Our oversight in prototyping that year was neglecting to consider that the interaction between the intake and cube went differently when either or both of them were on the platform surface or the ramp, so our lesson learned from that year was to include less than optimal arrangements of robots and game pieces in parameterization, and also make sure to test all the surfaces that we’ll encounter on the field. Also, we only had a wood platform and not the field official material (HDPE? I don’t recall what it was) which had a different interaction with the cover that was a bit of a surprise at our first event, so the following season we invested in the field specific material for early prototyping.

4 Likes

Not exactly the question you’re asking, but we had only 2 rollers per side which pretty much straight out normally, but we’re sprung so they could open wider and create a funnel, which worked quite well.

1 Like

Same here! Our springs were actually just Vex polycarb square tube. If you’re interested in dimensions, etc, they’re down in post #3:

For my team, the front wheel width was less about some sort of performance / grip advantage, and more about widening the effective area in which the robot could grab game pieces. I suspect that is what you’re seeing in these west coast intakes as well - my memory is that they used smaller wheels for packaging and to fit more contact points into the mechanism, so they made the outermost wheel wider to increase the intake’s range. The overall width seemed to be determined based on what would fit through their elevators.

Compliance in your intake and/or wheels should account for all cube orientations. I’d definitely recommend having some sort of system to “solve” diagonal cubes on the fly, which can be as simple as a passive curved piece of plastic at the back of the intake, or as complex as current monitoring on the intake wheels.

Can I suggest that the funnel was not the best 2018 intake design? We found that a single pair of wheels worked better, and did not depend on which way the cube was oriented (one side was shorter than the other).

The secret was that the “claw” holding the wheels was held “soft closed” by a reduced air pressure cylinder, such that it would open up as the wheels intaked the sides and corners of the cube. Since we used the same claw for lifting the cube, we used a separate solenoid valve to increase the pressure for a “hard grip”.

https://youtu.be/Coh4LwNu2Yc?t=27

2 Likes

It’s 2018.

You’re asked to intake boxes from the floor.

You scale from a 2015 design that works!

1 Like

I don’t think this was reflected in the reality of on field performance.

2 Likes

I disagree. Our robot intake outperformed most IMO.

1 Like

Outperforming most wasn’t your original claim.

Did it outperform the best angled intakes?

We’d have to do a lot of video analysis to confirm. I’m willing to be proven wrong, but yes that’s my impression. At minimum, our intake did not have to waste time righting tipped-over cubes, which as I understand it is something most of the “funnel” intakes (with fixed wheel spacing) had to do.

Yeah I think the main argument would be that they were so good at normal cubes that it made up for that.

1 Like

The best intake angle was 360 degrees

It was pretty good at scraping cubes away from walls

14 Likes

Though to the joke in the meme, looks like that Turing division full of green compliant wheels beat 5818, though the regionals went well.

This is true. The meme was made before we headed to Houston - as kind of a retort to the obvious design convergence that year. The real takeaway is to not be afraid to try really different things. This intake would happily consume a cube in any orientation. But, like many other designs, it needed many iterations and tweaks to work well.

Oh yeah. I’d be remiss if I didn’t repost this classic:

3 Likes

Since we’re talking about alternative 2018 intake designs now…

https://imgur.com/FamF7e3

3 Likes

Floptimus Prime, my beloved .

My favorite fact about that robot is that you guys used 11 full CIMs on it.

1 Like

The spinning flaps are what makes 5818 so iconic. I was deeply saddened upon discovering that your 2022 robot did not feature them. Michael, you would make me the happiest man in the world if you all brought the flaps back for 2023. #flapbacktoreality

1 Like