Perpetual Motion Machine?

Hello Forums,
I've been ruminating a bit over a design for a perpetual motion machine that utilizes buoyancy in order to function. Since such machines do not lose energy to their environment, there must be something flawed with my idea, but I'll post it anyway.

Here's a beautiful diagram I made in Paint:


Where did I go wrong?

Interestingly enough, I'm not sure where you went wrong.
As for how you would actually harness the energy; that's a different story.

I feel like that you wouldn't be able to keep the water-pressure high enough in order to retain the force needed to force the bouyant object down.

Allow me to clarify the design: the right half is filled with water, while the left half is filled with air. This is a side view of the system. Water lifts the buoyant object up on the right, and then drops it back down on the left. It is then picked up again by the water and the cycle perpetuates.

Allow me to clarify the design: the right half is filled with water, while the left half is filled with air. This is a side view of the system. Water lifts the buoyant object up on the right, and then drops it back down on the left. It is then picked up again by the water and the cycle perpetuates.

Aaaaah, makes more sense now; I was thinking something similar but not quite that.
I guess only issue would be preventing water from spilling into the other side yes? Otherwise it sounds like it would work just fine.

I went to the perpetual motion Wikipedia page and found this:

The "Float Belt". The yellow blocks indicate floaters. It was thought that the floaters would rise through the liquid and turn the belt. However pushing the floaters into the water at the bottom takes as much energy as the floating generates, and some energy is dissipated.

Picture:
[spoiler]

[/spoiler]

Seems very similar to my idea. I guess the area that makes this impossible is the transfer from air as a medium back to water at the gate.

That's a very interesting design that you found.
The belt design appears to help out with that gate flaw.
Otherwise, maybe some sort of forced pressure could also be used.