Skip to Content

Alan S. Tilles Comments on AMTRAK Train Crash and PTC Implementation

May 27, 2015

The unfortunate AMTRAK crash, May 2015, has created a surge of interest in a project that many of us have been working on for years, Positive Train Control.  Because of the media’s need to fill the 24 hour news cycle, there has been a glut of stories fingering blame in a number of directions and highlighting a number of issues.  As with many of these types of news stories, hype has led to misleading and inaccurate claims.  Stories have alternatively blamed Congress, the rail industry “dragging its feet”, lack of money, and lack of access to radio spectrum, to name a few.  In addition, there has been confusion on the difference between the old-technology of Active Train Control (ATC) and the new technology, Positive Train Control (PTC).  Let’s sort out the myths from the facts with regard to PTC.

The high profile Chatsworth train crash led Congress in 2008 to mandate the adoption of PTC by December of 2015.  An easy one page description of PTC can be found on the Union Pacific website. [1] To begin with, it’s important to remember that the original deadline isn’t until the end of this year.  Therefore, even if AMTRAK was farther along in its implementation of PTC, the implementation schedule would not necessarily mean that PTC would have been operational in this particular section of track prior to the accident.

PTC must be installed by freight and commuter railroads that meet certain criteria (too complicated to go into here).  AMTRAK is one of the commuter railroads that must meet the requirement (as are SEPTA, the Metro-North and Long Island Railroads, the Massachusetts Bay Transportation Authority and others).  Exempt railroads not required to implement PTC could certainly choose to do so.   In addition, this particular legislation relates to railroads, not subways.

Next, to meet regulations, the installed PTC system must be interoperable between all of the various railroads that share trackage.  AMTRAK (for example) must utilize a PTC system that is interoperable on the Northeast corridor with freight railroads like CSX, as well as multiple commuter railroads, including those mentioned above.

One criticism levied at Congress is that it did not allocate funding for PTC to be implemented.  Certainly PTC is an unfunded mandate for commuter railroads that are not for-profit entities.  Therefore, funding for this incredibly expensive project must be found somewhere.  The MTA (Metro-North and Long Island Railroads) estimate the cost of implementation, just for them, at over $968 million.[2] 

However, the lack of funding doesn’t tell the complete story, and there’s no guarantee that funding would have resulted in PTC installation already for AMTRAK.  Rather, there are a host of other problems and issues which make this system lengthy in implementation.

Since PTC must be interoperable across railroads, and the Federal Railroad Administration did not mandate all of the technologies or services for PTC, the freight and passenger railroads had to engage in discussions and come to an agreement on system parameters.  This creates a mix of for-profit entities and non-profit entities, each with their own priorities and restrictions.  That mixture is a prime recipe for delay.  Nevertheless, while the tempo of the process may have started slowly, that pace certainly has picked up steam over the past several years.

Remember, this technology has never been deployed before (although variations have been).  It represents an enormous (unfunded) undertaking by each railroad to create a nationwide PTC network.  Once those technical parameters are agreed upon, design must be undertaken.  And here lies an important issue.  Since PTC requires wireless spectrum to work, spectrum compatible for this purpose must be chosen and acquired.  In this case, the need for spectrum was not recognized by Congress in the original Act, and the Federal Communications Commission was not directed by Congress to make an allocation of useable spectrum.  Therefore, the railroads had an issue to face: where to find useable spectrum on a nationwide basis?  Since the system must be interoperable for railroads sharing trackage, spectrum selected for this purpose must also be available for every sharing railroad over a significant area.

The freight railroads had previously acquired spectrum in the 220-222 MHz band in most of the country, so they decided to repurpose this spectrum for PTC.  This spectrum also needed some technical rule waivers to be useable for PTC, too (furthering delays).  While it was a logical decision for the freights to repurpose spectrum which they already had, and not spend additional money for spectrum in another band, this resulted in commuter railroads having to acquire compatible spectrum, at least in those areas where the freight railroads do not have sufficient capacity to lease excess capacity to commuter railroads.  This created an additional conundrum – figuring out how much spectrum was necessary for the freights’ own systems.  This could not accurately be known until test systems were deployed, which typically cannot happen until after the system is designed (furthering delays).

With the knowledge that certain commuter railroads would need to bring their own spectrum to the table, a review of compatible spectrum was necessary.  Unfortunately, radio design[3] for PTC means that only three potential bands of spectrum could be used: (1) the aforementioned 220-222 MHz band; (2) the Automated Maritime Telecommunications Service (AMTS); and (3) the band formerly known as the Interactive Video Data Service (IVDS).  Each of these bands has its own problems for PTC, including (to various degrees):  different frequency channelization; technical restrictions that are incompatible with PTC (power, tower height, interference regulations, etc.); the lack of licensees from which the spectrum could be acquired (because many had failed to construct their initial licenses and turned them back to the FCC); and ongoing litigation amongst existing licensees.[4]

Recognizing this problem, we undertook a review in 2010 of potential options which would create dedicated, compatible spectrum for PTC.  We determined that one of the two blocks of IVDS spectrum could be reallocated by the FCC for PTC in every market where PTC spectrum was needed, except for New York City.  In other words, the FCC had at least one of the two IVDS licenses in its inventory (and in some markets both licenses), not licensed to anyone, in every relevant market, except for New York City.

It was the FCC’s intention to re-auction these licenses, for which there hasn’t been any significant use in decades.  In order to preserve that spectrum for PTC, in 2010 we filed a request with the FCC to reallocate one of the two IVDS blocks for PTC use.[5]  This would have solved the PTC spectrum issue for every commuter railroad (except in New York City).

The FCC is permitted to reassign spectrum designated for auction for public safety.  It was our contention that PTC is indeed public safety.  Unfortunately, other than receiving comments on the reallocation proposal, the FCC has never formally acted on our filing.  At the PTC Global Forum, one FCC employee stated that it was the FCC’s belief that it did not have the authority to reallocate spectrum designated for auction.   In a recent blog posting, the Chief of the FCC’s Wireless Telecommunications Bureau intimated something similar.[6]  It is apparently the FCC’s belief that Section 337 of the Communications Act enables the reallocation only where the applicant is a public safety eligible entity, not just where the service to be provided is public safety.  While we respectfully disagree, Congress could eliminate the PTC spectrum problem by simply directing the FCC to reallocate a portion of the IVDS band for PTC.

Correcting spectrum access now would be helpful for some railroads, but others have       already acquired spectrum (IVDS, AMTS or 220-222 MHz) from existing licensees.  Congress acting now on spectrum (as well as financial issues) would certainly be helpful for PTC implementation.  However, as discussed earlier, lack of ready access to spectrum for PTC cannot be seen as the sole reason for delayed implementation.  It’s only one of many factors, all of which should be addressed. 


[3]   Note that because the number of PTC radios that will be needed for implementation is not significant in number, many radio manufacturers passed on the opportunity. There aren’t enough sales possible for this limited use to make business sense for some radio manufacturers to invest in radio design and development.





Alan S. Tilles