In IEEE Spectrum's October 2018 issue, Ozan Tonguz - professor at Carnegie Mellon University, wrote about how Vehicle-to-Vehicle (V2V) communications could someday drive the ubiquitous tri-colored traffic light obsolete. (Online article here:
https://spectrum.ieee.org/transportation/infrastructure/how-vehicletovehicle-communication-could-replace-traffic-lights-and-shorten-commutes

Below is a YT video featuring a watered-down introduction to V2V:


The idea was to use V2V for a Virtual Traffic Light (VTL) algorithm, a system that will decide which car in an intersection would have the right of way. Aside from organizing traffic at crossroads, the solution aims to alleviate processor power consumption from machine-learned autonomous vehicles, as its simple procedure is less CPU expensive (remember IBM's token-passing in star-point topologies?).

While this solution seems fantastic at first glance, other parties could suffer drawbacks. Take pedestrians for example, who do not have antennas protruding over their heads to participate in V2V communications. They merely rely on the time the traffic light is red to get to the other side of the street. Unless there is an underground passage or overpass, pedestrians will be highly inconvenienced at getting to their destination (even if it is just a few blocks away). The article arguably suggests IoT as a solution, but it pretty much backtracks to the original problem where all the vehicles would need to stop (the VTL system can't acquire metrics on pedestrians - i.e. predicted walking direction and walking speed). 

How do you suppose these pedestrians would get across without traffic lights?


Even in the comments section of the online article, we find a few readers expressing their disapproval over this premature step.

"Shocking that anyone would be promoting such a bad idea that even they admit won’t work well with people walking or cycling. In fact, cycling and walking are the solution to traffic problems in cities."
-ReallySeriously2010

"This is somewhat frightening. As a cyclist or pedestrian do I really want to rely on "cheap dashboard-mounted cameras" to detect me? Were they a part of the simulation - how does this effect their commute times?

And what happens if I show up in a non-DSRC car to a retrofitted traffic light, do I just wait for the other DRSC cars to go?"
-Cliff Bargar

"The effects of this tech on pedestrians and cyclists is catastrophic and dystopian. After all, why stop at all? Why not subtly adjust spacing in real time to allow a constantly flowing weave? Call it a "virtual traffic circle." Everyone would love it but voila you've made it impossibly dangerous to step out your front door on foot, a right which had been untrammeled for centuries until auto companies invented "jaywalking"."
-Adam Bee

"This is dangerous not to say the least. First off, the pedestrians & cyclists. Secondly, driver distraction by constantly keeping their eyes on the dashboard instead of on the road waiting to see if a screen turns to red instead of green. Does the author of this article know how many road fatalities we have because of driver distraction within the automobile? common sense guys, common sense...."
-TobyS

Now, the concept of a virtual traffic light isn't all cons. The pros for drivers are plummy, as travel time on the road will be significantly reduced. However, I think there are still some tech. advances required before advantages are fully realized without any complications (as is the case with driver-less cars and electric vehicles).