This concept is not a new one and is something several transit agencies have attempted. Conceptually, a radio based detection system could be very accurate and provide reliable information in most applications. However, such a system must be more than reliable, it must be failsafe. Perhaps OPB can provide greater detail here as to why current radio based detection is not failsafe.

Much of this technology was originally developed for military applications. These systems are designed to operate reliably even in the presence of interference, but are not designed to be failsafe. Why? Several reasons come to mind. First, the military does not require their datalinks to be 100% available. They expect outages due to environmental effects, interference or other conditions. The general rule of thumb is a 95% availability. Additionally, while the military relies upon such datalinks, they allow for "acceptable losses", meaning they can accept a certain degree of failure from time to time. Finally, the cost involved in developing a failsafe, RF based system is cost prohibitive. This paradigm does not apply in the railroad world, where acceptable losses means zero losses. It would require a link availability of 100%, something which is not really practical - especially in where the system is being implemented in a subway (I cannot even begin to imagine the multipath and interference issues from that one). Additionally, the RF spectrum is becoming very crowded, adding another level of complexity. Nor is this a trivial matter, as businesses pay billions of dollars for spectrum allocation.

This is not to say it is not a good idea or should not be pursued, we just need to remember that the technology might appear compatible on the surface, it does not meet all the requirements for a failsafe implementation and will require a significant investment to develop a radio based detection system which is failsafe.