I'd much rather have this than "solar freaking roadways" .

you could try a "geothermal" approach.. underlay tubing through which you pump water.

Similar approach to geothermal pool-heaters.

You would only be able to harvest a modest temperature differential as compared to atmospheric temp (maybe 50C). 

That said, you could operate a stirling engine (better efficiency), peltier (worse efficiency), use for domestic heating or something funky (think "refrigerator that runs off of a cycling heat source"  http://apptechdesign.org/technologies/solar-refrigeration/ )

many people are playing with a geothermic design and to lay that much tubing is expensive. I want to make something that's cheap and easy.

To use the Seebeck effect you need hot and cold. The colder the better (more power). I see the heat but I don't see anywhere for the heat to go.

so if you think about it you have the heat in the asphalt and the cold of the ground underneath. Im hoping that will be enough to produce a good amount of electricity. I will do some experiments next weekend.

Some protection will be needed in case of street damages, so the remaining grid won´t be shorted.

If this works this will be the last thing we have to worry about I think. Making it cheap will be the hard part

Yes. A peltier solution would be surely too expensive. But a two-metal grid doesn´t cost that much. An asphalting machine could be modified to inlay metal stripes. Perhaps :-D I hope so!

yeah so we were talking about even a conductive paint that you could paint on old parking lots when they need to be resurfaced maybe applied with a robot so it makes the connections between the two kinds of paint? sooo many possibilities and things to try and develop. 

yeah so that would be the experimentation to see what can be produced with different methods of doing this. and yeah that I was going to do is make a bunch of squares of asphalt. 

I really like the idea. It would be awsome if you would build something like a square-foot of a street-like structure including the grid to estimate the power capability.

What are you best case expectations for  W/m^3??