Give up the fantasy that there is a large-scale demand for bikes, and stop wasting useful street real-estate on bike lanes. Austin is turning useful 4-lane roads (2 each way) into 2-lane roads (1 each way) with a nearly unused center lane and bike lanes that are used even less. Any time of day, winter and summer, you can see these empty, litter-collecting bike lanes doing nothing to help Austin's traffic problems. The UT area is the only place where bike lanes are actually useful and used.
Besides cutting down on car/truck usability, bike lanes give rise to a baroque series of restriping to create a maze of single-file lanes. That forces everyone to know which lane they need to be in far in advance and stay there, cutting down on maneuvering room if needed to avoid wrecks and possibly causing even more wrecks.
Has anyone ever stopped to consider what would happen if people DID use bikes to an extent that would make a perceptible difference in traffic? Imagine hundreds of bikers in the bike lanes every day. When all those bikes travel straight through an intersection, no cars can turn right; the entire right-turn lane has to wait until the bikes are all past, which would effectively make it impossible to turn if there are a significant number of bikes.
No one wants to bike to work -- much less run errands to the grocery store, hardware store, and dry-cleaners -- in 100º weather, or drizzle, or cold. The major users of bike lanes, aside from the UT area, are twenty- and thirty-somethings out for a Saturday-morning peloton. Other than serving as upper-middle-class exercise tracks, bike lanes are nothing but a shrine to that which people think SHOULD be used, but aren't actually used on a large scale, and won't be.
Prove me wrong. Make bike lanes contingent on use. Count the number of people riding in a bike lane, and if it's under an amount that significantly reduces car traffic (say, 100 bikes per hour, or 5% of the car traffic recently estimated at a major street), reclaim them for car and truck traffic.