I know it's impossible. I normally find the flaw behind the principle, but this one is hard to crack. The two wheels are geared together, so that no magnet on one wheel "backlashes" the wheel in a way that this magnets is between two magnets of the opposing wheel. I will use the term "opposing magnets" to describe magnets of two wheels facing one another and repelling one another.
The magnets on both wheels are oriented with the same pole pointing to the axis, so that opposing magnets repel each other. The net rotary force is zero, because the opposing magnets on the top side attept to spin each wheel in one direction while the magnet pair at the bottom apply the same force in the opposite direction.
The inventor believes that he can break this symetry by placing an iron plate between the lower portions of the two wheels. This plate is attracted to the magnets, and mitigates the repulsive force between two opposing magnets. The end result is (according to the inventor) that the opposing magnets on the top part have a stronger repulsion than the opposing ones at the lower part. This, is supposed to give a net non-zero force that should spin the wheels forever...
Where is the catch?
EDIT: This is not a v-gate engine or any type of single-wheel motor with a magnet mounted on a lever. The only moving parts are the two wheels, and the links in the duplicate question (those links which worked) show different designs. In addition, the link I provided is about a device whose dominant force is a repulsive, and not an attractive one. So, if the magnets lose their magnetic field, the device may act differently. Maybe the wheels will oscillate and then stop at a sticky point?