THE head-on collision between a Union Pacific freight train and a Metrolink commuter service on a single-track line in the Chatsworth district of Los Angeles on September 12 2008 offered a stark reminder of the harsh consequences of driver error. It was also an event that arguably changed US railway safety forever.
Assertions from leading figures from the Federal Railroad Administration (FRA) and National Transportation Safety Board (NTSB) in the immediate aftermath that Positive Train Control (PTC) could have prevented the accident, and the death of 25 people, spurred the federal government into immediate action. Almost overnight new legislation was drawn up by the Senate Commerce Committee and House Transportation and Infrastructure Committee stipulating that PTC must be installed by the end of 2015 on all main lines carrying passenger trains on a regular basis and lines that transport hazardous materials, which equates to an estimated 117,482km of track. President George W Bush subsequently signed the Railway Safety Improvement Act of 2008 into law on October 16 2008, leaving the US railway industry with just over seven years to get the show on the road.
PTC has existed in some form since 1982, and was listed among the NTSB's most wanted transport infrastructure projects during the 1990s. Take-up was generally slow, however, as railways balked at the high costs involved and its apparent limited safety benefits.
The 2008 mandate changed this. But with many of the older systems unable to meet today's essential requirements, new technology is needed, the rollout of which is proving to be an immense logistical challenge, and extremely expensive.
PTC systems must now be designed to prevent train-to-train collisions, derailments caused by excessive speed, unauthorised incursions on track sections undergoing maintenance, and movement of trains through incorrectly-set points. As a result the systems must be capable of determining the location and speed of trains, warn drivers of potential problems, and take action to stop a train if the driver does not respond to a warning.
Two dominant systems have since emerged that fulfil these requirements, predecessors of which are already in limited operation:
• Interoperable - Electronic Train Management System (I-ETMS), which is being adopted by the Class I freight and most commuter railways, and
• Advanced Civil Speed Enforcement System (ACSES) II, which is intended for higher-speed passenger operations, and will remain the primary standard for the Northeast Corridor.
Neither system is fully operational yet in the form required for introduction in 2015, with interoperability standards for both systems at different stages of development: some are already finalised while others are in draft form, and some have yet to be released.
Union Pacific, Norfolk Southern, BNSF, CSX, Canadian Pacific, Canadian National, and Kansas City Southern are working on solutions that meet the Interoperable Train Control (ITC) standards agreed in 2008 for use with Wabtec's I-ETMS system, the de facto standard Class I onboard system. ITC emphasises uniform interface standards, standardised message formats, wireless protocols, braking algorithms with interoperable hardware platforms, back office, wayside and base station equipment, and a locomotive platform that will enable infrastructure sharing. Development of new 220MHz data radios by Meteorcomm Communications (MCC), which is jointly owned by the "Big Four" Class Is, is another key element of the standardisation process.
A mix of pre-production and production hardware is available for the two systems from the "Big Six" PTC suppliers: Alstom, Ansaldo STS, GE Transportation, Invensys Rail, Siemens and Wabtec. They all emphasise the ease of installation and compatibility of their technologies with existing platforms.
Transportation Technology Centre Inc (TTCI) is at the forefront of PTC testing, developing software for its PTC testbed since 2005. More than two dozen engineers are currently working on two dozen projects for freight and commuter railways and federal agencies relating to PTC, utilising the facility's testbed comprising 80.5km of test track, a complete I-ETMS system and a V-PTC system, along with custom PTC test instruments. Among recent trials was the first over-the-air testing of several of MCC's pre-production ITC-compliant 220MHz PTC radios on trains, track-side units and base stations. The results were positive and mark a milestone in PTC development.
TTCI is also performing 220MHz RF coordination and radio network design in all of the large US urban areas for railways that will use the PTC-220 LLC spectrum. Simulation tools are in use to model all PTC message traffic loading for large-scale deployments for periods of up to several weeks.
TTCI says that its research and testing will impact PTC deployment in several ways, including reducing potential negative operational impacts of PTC through improving enforcement braking prediction algorithms, and by modelling and testing PTC system performance under various controlled stress conditions. TTCI's facilities and capabilities will also increase the efficiency, quality and quantity of testing and aid the completion of critical radio network designs for complex interoperable urban areas.
Nevertheless with such a tight timeframe for the development and deployment of PTC, the system that is rolled out might not be the optimal solution.
"Unfortunately this deadline does not allow time for implementation of more advanced PTC architectures like standalone systems that could achieve moving block or virtual block operation," says Mr Alan Polivka, assistant vice-president for communications and train control at TTCI. "Consequently the systems being implemented to meet the deadline are safety overlay systems that will generally degrade, not improve the efficiency of railway operations."
Polivka says the biggest concern with the deadline less than four years away is the operational impact of PTC. With functional and specification issues still being resolved, the system has undergone few performance tests in light-to-moderate operational scenarios, and none in dense operating environments. And with false enforcements and premature enforcement warnings commonplace in previous PTC systems, these are likely to continue to be a problem.
Past PTC systems have often had a negative impact on operating reliability. Overly-conservative braking algorithms, or loss of movement authority or signal status messages over the communications link which causes the PTC system to assume the most restrictive state, can result in trains slowing or stopping prematurely or unnecessarily.
"Since today's PTC implementations are overlay systems, they result in a significant increase in the amount of onboard, wayside and back office equipment, which introduces many more potential failure points that can stop trains," Polivka says. "In general, while potentially improving safety, PTC overlay systems have no means to improve operating reliability but many ways in which to degrade it."
Doubts also exist over PTC's effectiveness as a safety system on a network that already has a very good safety record.
The Association of American Railroads (AAR) estimates that only 4% of all train accidents on Class I mainlines are likely to have been prevented by a PTC system. And with an estimated $US 9.55bn cost to roll out the technology, which is expected to rise to an estimated $US 13.2bn as it is maintained over the next 20 years, the FRA estimates that PTC-related safety benefits will not exceed more than $US 674m, which equates to a $US 1 return for every $US 20 spent.
At Railway Interchange in Minneapolis last September delegates noted the steep financial burden that PTC is placing on railways, with Mr John Vogler, Amtrak director for communication and signalling maintenance, pointing out that while federal funds have helped, 85% of its capital budget is going to the expansion of its ACSES system.
The problem is extenuated by the FRA's ruling that PTC is installed on lines that carried hazardous materials in 2008 rather than 2015. With transport patterns changing significantly, the AAR estimates that railways will have to spend more than $US 500m to install PTC on around 16,000km of lines that will not transport passengers or hazardous materials in 2015. The association argues that costs could be reduced significantly by exempting lines that carry very few trains in these categories. A bill to amend this ruling, backed by the AAR, is currently being considered by the US Congress.
The federal government's stricken budget situation is similarly not helping the railways. Despite $US 50m being allocated for PTC in the 2010 fiscal year budget, $US 50m included in 2011 was eliminated in the compromise reached between House Republicans and Democrats during last year's budget crisis which almost led to the shutdown of the federal government. While this is only a drop in the ocean when it comes to the overall costs for railways, with further allocations unlikely, it demonstrates that the railway industry is largely going it alone.
So how might the Class Is and passenger operators get more bang for their PTC buck?
In a paper presented at Railway Interchange, Mr Steven Ditmeyer, adjunct professor in railway management at Michigan State University and a former associate administrator for policy and associate administrator for research and development at the FRA, points to research conducted by the AAR which states that railways are implementing PTC in a manner that produces minimal business benefits.
Rather than embracing this technology and improving their network operations, Ditmeyer argues that the railways seem to be embarking on a mission to prove that PTC offers no business benefits. With the FRA leaving the railways to develop and implement their own PTC strategy, he says they are hiding behind its huge cost-burden as a means to convince Congress that the law should be rescinded.
As a federally-mandated project, Ditmeyer says that railways should instead search for the optimal way to implement the technology. If PTC is done right, he believes it has the potential to reduce delays and costs as well as energy consumption and emissions while boosting capacity, reliability, customer satisfaction and safety. And with the ability to locate trains in real-time, PTC has the potential to revolutionise train control for many operators that currently only see the location of a train within a 32km window.
"There is a very similar argument surrounding the rollout of the next generation of air traffic control, where the established airlines are vehemently against it because they say there are no benefits," Ditmeyer says. "However, three smaller and bolder airlines are in favour of it because they can see how it will improve operations in the long term. Unfortunately too many in the railway industry see PTC as just another signalling system when it is a lot more than that, and has great potential to improve operations."
Ditmeyer argues that railways should adopt a network-centric approach to operations which is already used successfully by the military and the aviation industry. This involves implementing digital data communications, GPS positioning, sensors, and displays to serve the network, with the aim of delivering timely and accurate information to where it is needed, when it is needed, and to those who need it most.
Ditmeyer offers 13 recommendations on how this approach might be successfully implemented by railways rolling out PTC, ranging from specific technological improvements to revolutionising the management philosophy (see below).
He suggests a major hindrance to the deployment of PTC, particularly among the Class Is, is the failure to employ a system integrator to manage its rollout. He cites Southern California Regional Rail Authority's decision in 2010 to employ Parsons to manage Metrolink's PTC programme across its 824km network as the model that the Class Is and other railways should adopt.
Ditmeyer says that through this approach Metrolink is very open and clear with its staff and customers about the PTC rollout, what it is trying to achieve, and its impact. However, he feels that this is not the case for the Class Is where he says too many of its employees are unaware of its precise PTC plans, which he considers is a massive hindrance to an optimal rollout.
"PTC affects everyone in the railway," Ditmeyer says. "It alters the information flow throughout the railway, changing how it operates. Therefore everyone needs to know how PTC functions, and this includes the customers. Metrolink has done this with its passengers, who are informed that this programme is underway which is intended to prevent accidents like the one at Chatsworth in 2008. The Class Is need to do the same."
To overcome this, Ditmeyer says it is essential for the industry's senior figures to lead the deployment, and have a firm understanding of how PTC will affect the whole railway. This, he suggests, would immediately improve communication flows throughout the company leading to a more effective rollout because everyone - from software developers to unionised workers who will install the equipment - will understand what the end goal is.
Ditmeyer says this approach was adopted with encouraging results by Burlington Northern between 1987 and 1992 when its chairman led the company's Ares programme, an earlier form of PTC, developed by Rockwell International, which subsequently sold its technology to Wabtec. In this instance the entire railway was aware of the implementation, understood how it would benefit operations, with all entities working toward the same objective. However, a change in management in the early 1990s led executives to question its benefits. Ares was soon derailed and its implementation strategy died with it.
Ditmeyer says that unless US railways revert to something resembling this model, resistance to embracing new technology will always exist. He believes that PTC is a great opportunity but compares the industry's reluctance to embrace it with arguments against the replacement of steam traction with diesel traction in the 1940s and 50s. No-one questions that decision now, but while the industry continues to hide behind the cost-burden and endure with older technologies, the latest generation will never quite fulfil its potential.
13 ways to improve PTC deployment
1: Implement PTC systems that are integrated with other information systems such as Automatic Equipment Identification (AEI) systems, locomotive health monitoring systems, and precision dispatching systems. These offer accurate details of train length and weight and the status of dynamic brakes which can lead to the calculation of more accurate braking distances and improved understanding of where the ends of trains are located.
2: Improve flows of information by providing each employee in every department with all the information - and only the information - they need to do their job.
3: Keep data on trains, wagons, crews, and shipments secure and prevent unwarranted extraction of information from the digital data links communication network so that information remains genuine, unaltered and complete.
4: Install supervisory control and data acquisition (Scada) systems to monitor all wayside and onboard systems and components. These systems can report their status via the digital data link communications network to appropriate control centres and other organisations on the railway.
5: Resist the temptation to preserve investment in wayside signalling systems. Instead view such investments as sunk costs and move on. Understand that PTC does not follow the same logic as existing operating systems, and if this is followed, it will result in congested communication channels and reduced capacity and efficiency.
6: Hire staff with the right skills and who can be trained to work with the new system. Budgeting processes must be modernised to include benefits such as traffic increases from shorter transit times and improved reliability. This must be included in analyses, not just items like reduction in staff and fuel costs. Aim to optimise the PTC system as a whole, rather than specific subsystems.
7: Understand that organisational culture will change following the adoption of these changes. To be successful, telecommunication and train control staff will need to develop an extremely close working relationship.
8: The implementation team must represent all affected departments.
9: Recognise that network-centric operations will not solve all of the problems. It can lead to micromanagement. Use the information to improve the performance of the railway, not monitor the performance of individuals.
10: Top company leaders must understand the technology of PTC, establish the proper environment for its deployment, and provide real incentives for people installing it and those who will be operating it. The implementation team should be elevated in the company structure and be in regular communication with top leaders.
11: The implementation strategy must be clear and unambiguous.
12: The implementation strategy must be communicated throughout the railway to all departments, including suppliers and customers, and the unions who are key to a successful implementation.
13: Maintain realistic expectations. Thinking, planning, and organising are essential.