Getting to the point where most or all American drivers have safe and reliable driverless cars will take time. In the meantime, why don’t we have smart traffic lights or at least every traffic light operating with sensors?
There are several intersections near my house and work that clearly do not have sensors. You pull up in your vehicle and regardless of time of day or how many others want to go the same way as you do, you will wait a full light cycle. Some of these lights are one to one and a half minutes long. Sometimes this makes sense: one road clearly has more cars. Yet, often this full period passes with few to no cars going through the light.
But, let’s go further. I’ve also run into situations where the sensors might just be too sensitive. This occurs particularly between 9 PM and 5 AM when traffic is light on major roads. A single vehicle wanting to turn onto the major road can stop traffic. If you have a few of these on a single trip, this can be frustrating. Why not have coordinated light signals along major corridors? Cities and suburbs do not necessarily have to go to full-blown smart systems that hope to coordinate all traffic; even just doing this on a few main roads with significant amounts of traffic could help ease congestion.
Perhaps one issue is cost: what municipalities or other governments (depending on who has jurisdiction over the road) want to spend money on sensors and devices? One of the supposed benefits of driverless cars is that they will allow for more flowing traffic through coordination across vehicles. However, in that scenario the cost of less congestion is pushed to the car owners who have to purchase such a vehicle. Sensors at major intersections or at all intersections would not require anything or much from drivers. Yet, I bet you could make an argument that putting money into better intersections will be a cost savings in the long run with less time spent in traffic.