After much interaction -- both handholding and headbanging -- with the civil engineers responsible for the the Fremont Bridge, Seattle bicycle advocates have scored a victory, achieving significantly better bicycle accommodation on the bridge approaches than was planned a year ago. I've been personally doggedly determined on this one, and have closely observed, photographed, and thought through the bicycle, pedestrian, and car interactions on and around the bridge. I even took a day trip to Portland to photograph good bikeable bridges, and many good ideas from there have made it into the Fremont bridge. All of which homework was very helpful in making all the little cases and points that come up in a discussion of a complex multimodal environment.
At the outset, the engineers were bike-naive and even bike-hostile. (As a rule, traffic engineers don't speak bike; car is their native and only tongue.) But as a result of advocates' emails to the director and mayor, organized by multiple bike groups and sent by scores of dedicated bike advocates, we were able to pry open a dialogue. Then, by engaging in hours of discussion, comment, criticism (objective, constructive criticism, of the product, not of people), over the months, we were able to gain, bit by bit, a better bridge. By no means what bicyclists would have designed, but within the context of the flatly stated limitations of the project, way better than the less-than-mediocre status quo.
Several learning points: Traffic engineers can be stunningly uninformed, but after a number of months of close-in intervention, they and their product can improve. (They are, after all, civil servants, and they are rational, if a bit obdurate, human beings.) Traffic/street design is surprisingly unrigorous -- it's amazingly hunch-based and open to trial and error. Considering what's at stake -- safety, health, community, differential allocation of public space to different users -- decisions are made in a rather stultified way. But the good thing about such a non-process process is that it can be influenced, in spite of itself.