IMO variable compression ratio is the answer
 

Lets suppose your car has a 3.0L V6 and 10.0:1 compression ratio
then you have 300 cc that must be filled with air/gas mixture about an optimal 15:1 to ignite reliable and cleanly.

But during traffic jams or stoplights, it can switch to 30:1 compression ratio.
Thus you have only 100 cc to be filled with explosive mixture.

Of course an 3.0L engine at 30:1 compression ratio would guzzle more gas than a normal 1L engine at say 10:1 because of larger friction and pumping losses.

Do you think variable valve timing would be needed at all for significant consumption improvements?

Variable compression ratio - Wikipedia, the free encyclopedia

Variable Compression is ALREADY available by using a turbocharger, with the realtime adjustment of the "effective" compression-ratio occurring via computer-controlled dump-value!

It's a mechanically external "huff & puff" operation that 'recovers' some free energy from the exhaust...unlike complex, internal, VVT mechanisms which eat energy.

the answer is stop/start engine software and electric powered a/c, mild hybrids, etc.

also, in densely populated areas, mass transit/rapid transit.

I filed a patent application for a variable displacement-compression engine in 2004, but later it evolved into an infinitely variable in wheel drive with capacitive regeneration. The patent was issued in 2010.

regards
Mech

http://www.animatedengines.com/gnome.html

Yea variable compression is not the lowest hanging fruit, not even close. The practical limit for mechanical compression ratios is maybe 14:1, for full throttle. At say 20-30% load there is a benefit to increasing the compression ratio even if the valve timing mechanism allows minimal pumping loss, but why not just introduce a little (cooled) EGR to bump up combustion pressure/temperature?

Forced induction with a high mechanical "compression ratio" (we're really looking for the expansion ratio here), and say "Atkinson cycle" intake cams to increase low load efficiency and decrease detonation tendencies is essentially variable compression ratio.

I feel like the "upgrades" that engines are actually going to see on a large scale are direct injection (already happening), further upgradess to cam phasing systems (can get rid of most pumping losses if the range of movement for the cams is larger), forced induction (as a cheaper way to get most of the benefits of variable compression, variable displacement, and more complex variable valve timing), cooled EGR (already happening), and maybe laser ignition or something of that flavor. These things are highly reliable, simple, and relatively cheap to design in, compared to things like fancier valvetrains, variable compression, HCCI, etc. which have many more technical challenges.

Sounds like you want a diesel engine.

It's the expansion ratio that counts. If you could have 10:1 compression and 14:1 expansion ratios, that would be the key. Honda has a small engine that does this.

...and, Direct Gasoline Injection (GM's SIDI) allows the compression-cycle to happen WITHOUT any "bucking" counter-pressure occuring from advanced-ignition timing, because combustion ONLY occurs on the down-stroke/expansion cycle, when it's finally injected into the cylinder ala' diesel cycle operation.

Constant pressure (diesel) vs constant volume (spark ignition) combustion.

i think i read somewhere that someone has already planning for a variable compression car and has already developed a prototype. i think it was saab, not sure...