The patent application was initially rejected twice, and then on 10/1/2014, the attorneys added this language to claim 1:
wherein the initial track centerline data defines a centerline of the track, and wherein the subsequent track centerline data defines another centerline of the track.
and added this language to claim 33:
wherein the initial track parseable data comprises textual representations of the data in the initial database, and wherein the subsequent track parseable data comprises textual representations of the data in the verification database.
These additions were apparently enough to convince the US Patent Office that the invention was worthy of a patent. I haven't analyzed what these statements mean, but I've shown you where to look in order to understand why the patent was allowed.
Moreover, even though the application was initially rejected twice, the USPTO indicated that the original claims 27 and 28 were allowable from the get-go. Those claims originally read:
The system of claim 26, wherein the processing device is further configured to: compare the prediction data with the actual measured data for a plurality of trains at the specified portion of the track, thereby compiling expected performance data for the plurality of trains; establish an expected error margin value based upon the compiled expected performance data for the plurality of trains at the specified portion of the track; and log the identity of any train or locomotive where difference between the prediction data and the actual measured data exceeds the expected error margin value.
The system of claim 26, wherein the processing device is further configured to: compare the prediction data with the actual measured data for a plurality of trains at the specified portion of the track, thereby compiling expected performance data for the plurality of trains; establish an expected error margin value based upon the compiled expected performance data for the plurality of trains at the specified portion of the track; determine an acceptable error margin value associated with the expected error margin value; and log the identity of any train where difference between the prediction data and the actual measured data exceeds the acceptable error margin value.
and understanding why the patent was allowed would require analyzing the meaning of these claims (and interpreting them in light of the detailed description).
Just as a property line sets forth the boundaries of one's land, the claims set forth the boundaries of the invention. A patent is only broad if the claims are broad. This patent has lots of claims, and if you infringe just one claim, then you infringe the patent. But to infringe a claim, you must make, use, or sell a system that contains every feature described that claim. This is why a long claim is sometimes said to be easy to "design around." If your system omits just one feature required by the claim, then you no longer infringe that claim.
I found this info by navigating to the USPTO's site for public records, searching for patent number 9,002,545, and pulling up the "Image File Wrapper." From there, you can see the back and forth communication between the attorneys and the Patent Office.
I am no expert, and this response is for informational purposes only and not for the purpose of providing legal advice. You should contact your attorney or legal expert to obtain advice with respect to any particular issue or problem. Laws can differ dramatically from country to country, state to state, and technology field to technology field.