Okay, Jersey Mike, with regard to train braking performance with PTC, a few points:
1) Every train has a consist, and if it's not an accurate consist, it's potentially a violation of FRA regulations. And anyway, any wayside AEI reader will give you an accurate consist.
2) Weight of cars is known, because empty car weights are in UMLER (refer to Boylan post) and loads are listed on the waybill. This information is computerized, and has been available as a printout for the train conductor since at least the 1980s. Nobody has to enter a thing into the computer, because the information is already there.
3)Most railroads (probably all Class Is) mapped their lines using GPS some years ago. So a complete three-D route profile is available: grade and curvature, plus speed limits, grade crossings, signal locations, bridges, tunnels, etc.
4) Since the on-board computer uses GPS location (in addition to odometer wheels and fiber-optic gyroscopes (very cheap), it knows exactly where the head end of the train is, and since it has consist detail, it also knows not only the train's length but its precise position on the railroad. It can actually handle a full service application as well as an experienced engineer. I saw it done, up on Minnesota's Iron Range, with a 22,000 ton, 180 car taconite train -- in 1988.
5) The 1988 demonstration was with 8087 processors, a 4800 baud data link, and limited computer memory. Computers have improved A LOT since then. But BN's ARES worked just fine. BN didn't implement it systemwide because there were those who wanted to wait for the hardware and software to get better and cheaper -- and they have.
BN concluded the (successful) ARES demonstration in 1991. That's almost two decades ago. What have the railroads been waiting for, an engraved invitation (or Federal money)?
Randy Resor, aka "NellieBly" passed away on November 1, 2013. We honor his memory and his devotion to railroading at railroad.net.