I do not have any material to post on this, but, in the late 1940's, Union Switch and Signal did some preliminary work on developing a fixed signal system for airports that would have governed use of runways and taxiways by signal indication, in a manner similar to CTC governing the movements of trains. There was no interest, and it died on the vine. There may be some old records of this on some airliner or FAA website.
Wyhog, the only thing I can think of that makes railroad signals so expensive is the exposure to tort liability and the elaborate interaction of track circuits (all installed, tested, and maintained manually). Everything has to work as intended every time in any weather, and be tamper-proof and protected from lightening damage. A highway traffic signal controls one intersection and is usually not linked to adjacent intersections, does not detect occupancy or the integrity of the road between signals, and usually only detects approaching vehicles by a non-fail safe system such as movement detectors. The only thing that has to be fail-safe on a traffic signal is that it must not allow two interesting routes to proceed at the same time. Railroad signals are usually linked to: two or more adjacent blocks in each direction; all the turnouts and some derails within their track circuits; some road crossing warning devices; and they require batteries, commercial power, and a lot of digging. Whenever we did any track project, the signal cost was usually more than half the total cost. And the Signal Department would not negotiate the price at all.
Somebody told me that, when the Milwaukee Road built the new Milwaukee, Wisconsin, station in the 1960's, they used Eagle highway traffic lights at the depot. Perhaps a Forumite can confirm or deny this and also tell us whether they were used as actual block signals or merely as switch indicators?