How should Ben continue building his automatic can crusher? Post your idea + you might get a free Ben Heck t-shirt!

I watched a little of the video.

What significance to physics is there that the building at CERN is the same shape as a magnet that makes pretty plasma shapes when placed in a bell jar of low pressure hydrogen?

It is misdirecting the viewer by finding a shape in a building at place of known scientific research and presenting it as significant when the same shape can be found everywhere.

Space hoppers, The Epcot centre and soap bubbles all have shapes that would no doubt make similar pretty patterns of plasma in a similar situation.

Always be suspicious of Youtube videos with comments disabled. :-)

I just rewatched the can crushing episode and had a few thoughts on how to better achieve the goal.

The first thought I had was to do with the strength of the sides of the cans.  With the sides of the cans being perfectly straight it takes quite a bit of force to crush a can.  Push on both ends of a can with your hands, and it's very difficult to start the crush.  Put some dents in the sides and it crumbles much easier.  Putting dents in the side of the cans can be hard to automate and can also result in some very unevenly crushed cans as it tends to fold in the direction of the biggest dent.  Now if you take another can and grip both ends of it and twist them opposite ways while pushing in, it is much easier to crush the can because the twisting motion causes the sides to dent inwards in a very even pattern.  To automate this, you simple need to make the surface the can is crushed against spin and covered with rubber or some other high traction material and have the end of the crush piston be covered in the same material to keep the whole can from spinning when it hits the spinning plate.  You may also want to key the piston to the crush tube to keep it from rotating.

My second thought was with the actual force used to crush the can. While your stepper motor was able to provide the force needed to crush the can, it did so very slowly.  You were very concerned with the direct force applied to the can from the lever/piston from a complete stop.  I don't think that this is the way to go.  If you go slow enough a midsized person can stand on a can, so they can hold a fair amount of weight before they start to collapse, but if you lift your foot up and move it down quickly it doesn't take much effort to crush it.  F=MA is the key here, if the A is small, then it takes much more M to cause the same F.  I propose that if you were to use your final piston idea, give it a more mass and moved it fast, it would provide much more force than a slow moving lightweight piston.

A fast heavy piston combined with the spinning flywheel clutch of doom from my first thought should provide a fast can crushing rig.  Hardest part left would be automating the loading/unloading so that the machine could actually be efficient.  I'm envisioning something similar to an upside down machine gun that uses the motion of the piston to load the next can into the crush chamber, but this would require a long throw on the piston and very tight tolerances.