Is being a “rude” driver better for traffic?
Source: https://www.nytimes.com/2016/10/13/us/why-last-second-lane-mergers-are-good-for-traffic.html
If you’ve ever merged into a highway, you might think that merging as early as possible is the safest and nicest option to do. Merging early removes the possibility of not being able to merge at the end of the road, and isn’t as rude as cutting off people in front of you. So why fix what’s broken? It turns out that merging as late as possible is best for traffic and driver safety, and this idea is known as the “zipper merge.” If you think about how the two sides of a zipper zip up into one side, you can also imagine cars from two lanes merging into one line in a similar fashion, by alternating which car goes at a time at the merge point. This actually is the more efficient way of merging two lanes, as it uses both lanes as much as possible, and makes it clear when a car will actually merge. Some work zone engineers talk about seeing one lane not moving at all and then the adjacent lane moving fast and then merging into the slow lane. However, after zipper merges were introduced, the engineers commented that the two lanes were moving much more smoothly, and both lanes were much shorter. From a purely practical point, it definitely makes sense why this would be the most efficient, but there definitely are other mental factors, such as perceived rudeness of the zipper merge. I’m sure that you’ve probably experienced somebody cutting you in line for something before, and the zipper merge is in a sense “cutting in line.” Many drivers still aren’t willing to change to the zipper merge, and thus may try to block off people attempting to merge, causing huge safety hazards. However, research reports a 15% increase in traffic flow and a 50% decrease in lane length through usage of the zipper merge, and the benefits of the zipper merge definitely cannot be denied.
I found this article really interesting, because as a driver, I always thought that it was best to merge as early as possible to be safe. However, after reading this, it definitely makes sense why zipper merging would be more efficient due to the alternation of drivers when merging. I also think that having zipper merge prevents “phantom traffic,” which is the idea that cars that suddenly slow down cause traffic. By having zipper merge, cars won’t unexpectedly slow down, because they will be expecting alternation of merging at the merge point. However, without zipper merge, people can just randomly merge into lanes, causing sudden stops which leads to more traffic. However, I think the shift to zipper merge will be difficult, because most drivers are accustomed to merging early, and is defensive approach to driving, something emphasized heavily in driving courses. I, personally, would rather merge early and ensure that I’m able to merge rather than wait until the end and then try to merge, despite the benefits of zipper merging.
But why is it that zipper merging is optimal? How is it that the “greedy” approach of racing to the front of the lane before merging actually is the optimal solution? This is actually the idea of Braess’s paradox! If we give drivers more choices of locations (analogous to adding more roads to ease traffic) to merge, then it will cause traffic to be more congested. With zipper merge, the network would look something like this:
The big dot (node) represents the possible merge location, and the 3 edges represent lanes. On the contrary, without zipper merge, the network would look something like this:
The dots (nodes) represent possible merge locations, and the edges represent possible routes to take (diagonal lines are the path to merge into the other lane). We can see that without the zipper merge, there are more possible routes to take (more roads introduced), and thus it definitely makes sense why it could lead to more congested traffic.
Overall, I find the idea of zipper merging fascinating, and I think it provides a very solid structure for merging lanes, as it would clear any confusion as to when people should merge (early vs late). I definitely am excited to see the receptiveness of drivers to zipper merging, as many states are currently advocating for zipper merging, and I think approaching the idea of lane merging (and in general traffic) with networks is extremely fascinating!
interesting!