Per the DOT, the average age of cars on the road in the U.S. is a record high 11.5 years. While swarm communications would undoubtedly be benficial with much less than 100% of the cars on the road participating, we are a long ways off from even 10 or 15% of cars having it.
The interesting thing about swarm/train technology is that it actually allows a lot more cars to move along a lane at the same time (and with less energy.) Autopilot already helps some by mitigating the too close/overreacting driver pileups, but swarm technology would allow autonomous cars to get much closer together safely.
In places with lots of congestion and HOV lanes to try to mitigate the congestion, I'm thinking that "swarm" or "train" lanes will show up while there still aren't that many cars as a fraction of the total population. These will be much more effective than HOV lanes at actually reducing congestion provided there are enough cars able to use them.
Having those lanes will in turn greatly increase demand for cars capable of using them among people who are frustrated with long commutes. Even taking a few percent of the cars out of the traffic jam substantially reduces the jam's size and duration.
Tesla's ability to do OTA software updates and inclusion of AP hardware on all cars (since September 2014) makes them uniquely positioned to deliver a massive fleet of swarm capable cars in very short order, assuming one of the existing radio interfaces can be adapted to the desired purpose - and the car already has most of the current interface standards onboard (cellular, WiFi, Bluetooth.) Even if they need to add a communications module, it's probably a rather simple, inexpensive retrofit.
I haven't read of a proposal for a swarm standard (I'm sure there are some) - you need very low latency, moderate range (200m?) but don't need terribly much bandwidth to make it practical.
Walter
