MxV Rail A subsidiary of the Association of American Railroads Pueblo, Colorado USA www.mxvrail.com Technology Scan Summary Report for Advancements in Railway Research for Transport Canada Prepared by MxV Rail P-24-004 April 02, 2024 Disclaimer: This report was prepared for Transport Canada (TC) by Transportation Technology Center, Inc. (MxV Rail), a subsidiary of the Association of American Railroads, Pueblo, Colorado. It is based on investigations and tests conducted by MxV Rail with the direct participation of TC to criteria approved by them. The contents of this report imply no endorsements whatsoever by MxV Rail of products, services, or procedures, nor are they intended to suggest the applicability of the test results under circumstances other than those described in this report. The results and findings contained in this report are the sole property of TC. They may not be released by anyone to any party other than TC without the written permission of TC. MxV Rail is not a source of information with respect to these tests, nor is it a source of copies of this report. MxV Rail makes no representations or warranties, either expressed or implied, with respect to this report or its contents. MxV Rail assumes no liability to anyone for special, collateral, exemplary, indirect, incidental, consequential, or any other kind of damages resulting from the use or application of this report or its contents. i Acknowledgements MxV Rail Project Manager Christopher Pinney Subject Matter Experts Mr. Benjamin Bakkum Dr. Yin Gao Dr. Silvia Galvan-Nunez Dr. Duane Otter Mr. Alan Polivka Dr. Anish Poudel Mr. Brian Smith Mr. Tony Sultana Dr. Stephan Wilk Dr. Matthew Witte Editorial Services Susanna Tobias Transport Canada Project Manager William Anthony i Table of Contents 1.0 Introduction .........................................................................................................................6 2.0 Research Themes Summaries ............................................................................................6 2.1 Track and Infrastructure Inspection Technologies ................................................... 7 2.1.1 Current Landscape ......................................................................................7 2.1.2 Opportunities for Future Research ..............................................................8 2.2 Equipment and Rolling Stock Inspection Technologies ......................................... 11 2.2.1 Current Landscape ....................................................................................11 2.2.2 Opportunities for Future Research ............................................................11 2.3 Fire Risk Monitoring and Mitigation Technologies .................................................. 13 2.3.1 Current Landscape ....................................................................................13 2.3.2 Opportunities for Future Research ............................................................14 2.4 Flooding/washout Risk Monitoring and Mitigation Technologies ............................ 17 2.4.1 Current Landscape ....................................................................................17 2.4.2 Opportunities for Future Research ............................................................17 2.5 Geohazard Monitoring/Mitigation Technologies ..................................................... 19 2.5.1 Current Landscape ....................................................................................19 2.5.2 Opportunities for Future Research ............................................................19 2.6 Technologies for Managing Cold Weather Operational Challenges ...................... 21 2.6.1 Current Landscape ....................................................................................21 2.6.2 Opportunities for Future Research ............................................................22 2.7 Technologies for Subgrade Stabilization ................................................................ 25 2.7.1 Current Landscape ....................................................................................25 2.7.2 Opportunities for Future Research ............................................................25 2.8 Enhanced Train Control ......................................................................................... 27 2.8.1 Current Landscape ....................................................................................27 2.8.2 Opportunities for Future Research ............................................................28 2.9 Technologies for Increasing Grade-Crossing Safety .............................................. 31 2.9.1 Current Landscape ....................................................................................31 2.9.2 Opportunities for Future Research ............................................................31 2.10 Technologies for Monitoring Trespassing .............................................................. 33 2.10.1 Current Landscape ....................................................................................33 2.10.2 Opportunities for Future Research ............................................................34 2.11 Development of Tools and Analytics for Risk Assessment .................................... 35 2.11.1 Current Landscape ....................................................................................35 2.11.2 Opportunities for Future Research ............................................................35 2.12 Cybersecurity ......................................................................................................... 37 2.12.1 Current Landscape ....................................................................................37 2.12.2 Opportunities for Future Research ............................................................38 2.13 Human Factors ....................................................................................................... 40 ii 2.13.1 Current Landscape ....................................................................................40 2.13.2 Opportunities for Future Research ............................................................40 2.14 Low Carbon Fuels .................................................................................................. 42 2.14.1 Current Landscape ....................................................................................42 2.14.2 Opportunities for Future Research ............................................................42 2.15 Zero-Emission Rail Propulsion Technologies ......................................................... 44 2.15.1 Current Landscape ....................................................................................44 2.15.2 Opportunities for Future Research ............................................................44 2.16 Carbon Life Cycle of Track Materials and Rolling Stock ........................................ 47 2.16.1 Current Landscape ....................................................................................47 2.16.2 Opportunities for Future Research ............................................................47 2.17 Decarbonization of Rail Maintenance Activities ..................................................... 49 2.17.1 Current Landscape ....................................................................................49 2.17.2 Opportunities for Future Research ............................................................49 2.18 Alternative Fuel and Battery Tender Cars .............................................................. 51 2.18.1 Current Landscape ....................................................................................51 2.18.2 Opportunities for Future Research ............................................................51 Appendix A: Research Themes ................................................................................................ A-1 Appendix B: Industry Stakeholder Interviews ........................................................................... B-3 Appendix C: Reference Materials ............................................................................................. C-9 iii List of Figures Figure 1. CPKC hydrogen fuel cell locomotive .............................................................................6 iv List of Abbreviations and Acronyms Abbreviation/ Acronym Meaning AC Alternating current ACO Ant Colony Optimization AAR Association of American Railroads AI Artificial intelligence ALTRIOS Advanced Locomotive Technology and Rail Infrastructure Optimization System ANFIS Adaptive neuro-fuzzy inference systems ANSI American National Standards Institute ATCS Advanced train control systems ATI Automated track inspection ATO Automated Train Operation BEL Battery-electric locomotive CARE Community, Analysis, Response, and Evaluation CBTC Communications-based train control CFD-DEM Computational fluid dynamics-discrete element method CNG Compressed natural gas CNN Convolutional neural network CPKC Canadian Pacific Kansas City CTA Cognitive Task Analyses CWR Continuous-welded rail DAS Distributed Acoustic Sensing DoS Denial of Service DOT Department of Transportation DTL Diode–transistor logic ETCS European Train Control System FHWA Federal Highway Administration FRA Federal Railroad Administration EIC-PRT Employee-in-Charge Portable Remote Terminal EMS Energy management systems ERP Effective radiated power ETC Enhanced Train Control EO-PTC Enhanced Overlay PTC FMB Full Moving Block GA Genetic Algorithm GHG Greenhouse gas GNSS-RTK Global Navigation Satellite System-Real-time kinetic positioning GPR Ground Penetrating Railroad HBD Hot box detectors HEP Human error probability HIPS Host intrusion prevention systems HMI Human-machine interface HRCTC Higher Reliability and Capacity Train Control ICD Interface Control Documents InSAR Interferometric Synthetic Aperture Radar ITC Interoperable Train Control LAN Local area network LDS Location determination system LiDAR Light Detection and Ranging LNG Liquified Natural Gas LP Liquid propane v LRV Light rail vehicle MITM Man-in-the-middle MOTES Mobile Telematic Systems NCHRP National Cooperative Highway Research Program NDE Nondestructive evaluation NRC Nuclear Regulatory Commission NREL National Renewable Energy Laboratory NTSB National Transportation Safety Board OBRD Onboard Broken Rail Detection OEM Original equipment manufacturers PSF Performance shaping factors PSO Particle Swarm Optimization PTC Positive Train Control PTL Positive Train Location QMB Quasi-Moving Block QRA Quantitative risk assessment RF Radio frequency RGHRP Railway Ground Hazard Research Program RIoT Rail Internet of Things RISC Rail Information Security Committee ROI Return on investment SEPTA Southeastern Pennsylvania Transportation Authority SME Subject matter expert SPAR-H Standardized Plant Analysis Risk-Human TADS Trackside acoustic detector systems TC-IC Transport Canada’s Innovation Centre TG Track geometry THD Truck Hunting Detectors TPD Truck Performance Detectors UAV Unmanned aerial vehicle UP Union Pacific WILD Wheel impact load detectors WPD Wheel Profile Detector 6 1.0 INTRODUCTION In late 2023, Transport Canada’s Innovation Centre (TC-IC) requested MxV Rail’s support in developing a technology scan of emerging railway technologies. Designed to focus on technological advancements in long-haul freight and inter-city passenger rail, this technology scan included a review of technology references from around the world and Canadian industry stakeholder views on technology trends and important research opportunities. The objective of the technology scan is to provide information to the Canadian Rail Research Advisory Board and support the development of TC-IC 2025–28 Research, Development, and Deployment (RD&D) Work Plan. Figure 1 illustrates an emerging technology (hydrogen fuel cell locomotive) identified in the technology scan as being developed to address railway decarbonization strategies. Figure 1. CPKC hydrogen fuel cell locomotive 2.0 RESEARCH THEMES SUMMARIES Transport Canada and MxV Rail SMEs scanned technology references from around the world and gathered industry stakeholder perspectives from Canadian Class I railroads, short line railroads, and academic research institutions based on eighteen research themes representing emerging technologies in the railway industry (see Appendix A). Using SME knowledge, information from the Canadian stakeholder interviews (see Appendix B), and technology scan references (see Appendix C), the research themes have been developed into individual summaries, each containing a current landscape and opportunities for future research section. 7 2.1 Track and Infrastructure Inspection Technologies 2.1.1 Current Landscape Ensuring the safety of railway infrastructure is crucial, and regular inspections are necessary to monitor and detect anomalies that may lead to damage. Researchers are considering a variety of inspection techniques for conducting track inspections at different development stages. These techniques include ultrasound, phased array ultrasound, electromagnetics, thermography, machine vision, 3D Light Detection and Ranging (LiDAR), photogrammetry, unmanned aerial vehicles (UAVs), Drone-Based Digital Image Correlation, Distributed Acoustic Sensing (DAS), and track geometry measurement systems that use lasers, accelerometers, cameras. Despite the development of computer-aided track inspection technologies, manual inspections continue to be the current predominant inspection method, especially in the search for missing or broken track components. Despite being the main inspection method, manual inspections can be both subjective and challenging due to the logistics of performing detailed inspections at speed (revenue service or hi-rail). Typically, human inspectors conduct visual inspections and document any defects found through field measurements and visual observations. The data is usually recorded in paper-based forms that must be well-organized to support future inspections. In addition, such processes may be limited by 1) the experience and knowledge of the inspector and 2) the location of the area to be inspected, especially for larger structures like railway bridges. Recently, machine-based and automated track inspections (ATIs) have been explored to augment human inspections. Nondestructive evaluation (NDE) techniques have been widely implemented in rail and wheel inspections, and these techniques continue to improve in detecting material failure in track components. Additionally, ATIs for track geometry (TG) measurements make it possible to improve track inspection in a more efficient and effective manner. Railroads seek to increase automated TG measurements and reduce periodic visual track inspections. In addition, researchers have investigated using machine vision inspections (e.g., drone-based imaging and image processing technologies) and providing consistent, automated, and less- biased inspections. By leveraging artificial intelligence (AI), machine learning, and deep learning approaches, such as convolutional neural networks (CNNs), researchers are trying to develop a fast, accurate, easily operated, and low-computation track inspection method that uses imaging technology. In addition, the use of the digital twin concept is also being considered for the complete management of railway infrastructure, which is believed to offer continual asset monitoring and life cycle assessment by incorporating sensors and inspection information into the digital twin. These technologies present several opportunities for efficient and effective track inspections. Some of the relatively mature technologies used for monitoring track infrastructure, such as NDE, continue to evolve, while other technologies (ATI, AI-based, digital twin) are still in the early stages of development. Overall, the inspection technologies present new horizons and opportunities for future research that can help achieve accident-free rail transportation. 8 2.1.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Development, testing, and validation of advanced NDE techniques, such as phased array ultrasound, laser ultrasound, infrared thermography, etc. Rail inspections for detecting and characterizing internal anomalies are primarily conducted using a traditional ultrasonic testing method. Advanced NDE methods present newer horizons and opportunities for future research that can help achieve accident-free rail transportation. The research outcome can provide information on the performance of advanced NDE techniques and shed light on their limitations in revenue service implementation. The outcome of the research can inform the reliability of the new NDE methods for rail inspections in real- world scenarios. Enhanced rail inspection techniques and procedures using appropriate tools can lead to enhanced rail safety. The research can equip stakeholders, in particular railways, with strong diagnosis tools to both detect internal faults and prevent accidents Integration of other NDE technologies such as ultrasonic, radiographic, and optometric, with automated track geometry measurement system (ATGMS) There are a few research projects that have been conducted on ATGMS and its capabilities regarding automated fault detection in rails, but there is a lack in research in evaluating the use of multiple technologies for optimal fault diagnosis. The outcome of this research can inform the optimal sets of technologies that can be integrated with ATGMS, which enables identifying faults that advanced track geometry measurement systems cannot detect single handed. Optimal autonomous inspection system enhances rail safety and provides scientific evidence for supporting regulatory bodies and policy makers. The tool can help railways enhance supply chain resiliency by planning for corrective maintenance. Development, testing, validation, and improvement of drone- based monitoring technologies for track and infrastructure inspection There are research projects focused on rail infrastructure using drones with optical machine-vision cameras and LiDAR for mapping rail infrastructure objects and surfaces. These technologies use image processing to detect rail flaws and broken tracks. This research informs the best practices for flying over rail infrastructure, setting up automated mapping, and developing stabilization and image enhancement methods. Another value proposition of this research is improved data analysis to perform optimal fault diagnosis and capture rail faults. This tool has the potential to assist rail inspectors with rail and infrastructure inspection. The tool can also help railways to support effective and reliable track maintenance Implementation and enhancement of remote rail track monitoring technologies using distributed acoustic sensing in revenue service environment. Literature suggests acoustic sensing based on fiber-optic technology along railroad tracks allows for anomaly detection. However, further research is required to The outcome of this research can 1) shed light on the performance of acoustic sensing and fiber-optics technology in a revenue service environment and 2) The technology can be used by railways as a fully remote anomaly detection tool to both support effective and reliable track maintenance and 9 enhance the performance of such a system with the focus on improving signal-to- noise ratio and the reliability of technology in revenue service-type environments. develop signal and data analysis processing methods that assist in detecting anomalies. reduce the number of incidents. Development of training modules for railway personnel, conductors, and inspectors to use and leverage AI-based failure diagnosis technologies. There are several inspection technologies that can leverage AI to identify faults along rail tracks. However, there is a lack of training modules or learning materials to instruct railway personnel and rail inspectors on the use of these tools. The outcome of this research is development of training modules for railway personnel and inspectors on the efficient use of inspection tools that leverage AI. The training modules help rail personnel effectively use AI- based track inspection technologies, leading to enhanced rail safety and fewer incidents. The training modules can help rail inspectors with enhanced and smoother inspection procedures. Optimizing and fine- tuning the AI models for broader applications in railroad condition monitoring, considering different types of infrastructure and environmental conditions. There are several current research projects investigating the performance of AI- based tools in track inspection, but very few projects considered different environmental and infrastructure conditions. The outcome of this research is the creation of tools that are designed for a broader range of environmental conditions and that can effectively perform a safety inspection. The research can help railways identify track faults in different environmental conditions leading to safer operations and fewer accidents with regard to track faults. Research on smart trains with multiple onboard sensors and data fusion. There are several research projects investigating the use of on-board sensors, such as Sonar, machine- vision cameras, air- coupled ultrasonic transducers, and accelerometer- coupled AI/ML algorithms. These concepts are only demonstrated in controlled laboratory settings or on smaller track sections. The proposed concept can potentially be operated at revenue speeds, with real-time data acquisition and processing capabilities, highlighting the concept of a “smart train.” The research can help railways identify track faults in different environmental conditions, thereby leading to safer operations and fewer accidents with regard to track faults Onboard broken rail detection using electromagnetics. The concept was tested in both laboratory and field environments. The test results proved the viability of Onboard Broken Rail Detection (OBRD) with the transmission coil successfully inducing detectable signal in the The outcome of this work can provide a valuable understanding on how to eliminate crosstalk between Tx and Rx coils, how to add frequency selectivity to the shunts in the test setups, carrying out track Onboard broken rail detection using electromagnetics presents unique advantages to PTC and or CBTC systems in identifying rail breaks. 10 rail. Additionally, data analysis of alternating current (AC) track impedance was conducted to provide insight on trends for impedance variables with varying frequencies and under different environmental and ballast conditions. impedance characterization over a longer section of track, and developing a breadboard and prototype of the system for testing. 11 2.2 Equipment and Rolling Stock Inspection Technologies 2.2.1 Current Landscape Mature wayside inspection technologies include wheel impact load detectors (WILD), Truck Performance Detectors (TPD), Truck Hunting Detectors (THD), Wheel Profile Monitors (WPD), hot box detectors (HBD), acoustic detector systems (TADS), and hot and cold wheel detectors. These technologies are mature in the sense that the systems are commercially available and installed in great numbers throughout the North American rail network. All are monitored through the Association of American Railroads (AAR) committee structure with shareable data centralized through Railinc. Commercial contracts and intellectual property agreements limit the data details that can be shared, but there is still a substantial amount of information being warehoused. Research to address the systems interdependencies between the data available from different detectors (e.g., how wheel profile wear per railcar correlates with truck performance detector data) would likely add intelligence to maintenance practices, and ultimately, to safety. Emerging technologies are those technologies that either have limited commercial applications or are still in the development stages. Examples of emerging technologies would include machine vision systems capable of imaging the entire railcar at track speed. Full-train inspection portals are showing up on the network, but the software analytics for fully automated inspection are still evolving. Thermal vision inspection systems for locomotives are also under development. BNSF railroad is leading much of this development. The concept of full-train thermal imaging is an area ripe with research opportunities. Cracked wheel and cracked axle detection are being actively researched both in North America and worldwide. For example, China is using piezo-based ultrasonic automated wheel inspection systems. This technology was evaluated for use in North America, and the evaluation led to research using electromagnetic acoustic transducers. On-board sensing technology is in its infancy in North America and requires much additional research. A commercial consortium known as RailPulse is now monitoring equipment with on-board sensors operating on the network. The AAR is sponsoring the ongoing development of Mobile Telematic Systems (MOTES) through a technology advisory group with representatives from the industry and supply community. 2.2.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Software and analytics development to improve the utility of the data. Historical data is accumulating in large volumes across the industry. The application of machine learning techniques to analyze and extract insights from historical rail data. Advanced data analytics tools for rail operations, predictive maintenance algorithms, and decision-support systems to optimize rail network performance. Providing tools to analyze and use historical data effectively, leading to improved decision making, operational efficiency, and predictive maintenance capabilities. Conventional analytic approaches can be applied for data mining The use of data mining techniques to identify trends in rail operations Enhanced AI algorithms for machine vision inspection Improving the accuracy and efficiency of rail inspection systems, 12 to reveal trends and for training AI algorithms. AI is evolving rapidly across the industry and is being applied to improve the accuracy of existing machine vision inspection systems. Much of the performance research in this field is proprietary and closely guarded by the owners for competitive purposes. and the development of AI-powered predictive maintenance models. systems, data mining tools to uncover trends in rail operations, and improved predictive maintenance models for rail infrastructure. enabling predictive maintenance, and enhancing the overall safety and reliability of rail operations. Development of validation and calibration procedures for wayside and on- board sensors. This leads into the larger issue of standardization and regulation of automated inspection systems across north America. Development of sensor calibration methods for railway applications, the standardization of automated inspection systems, and the regulatory framework for deploying these systems. Standardized procedures for sensor calibration, regulatory guidelines for automated inspection systems, and improved wayside and on-board sensing technologies for rail safety and maintenance. Enhancing the accuracy and reliability of sensor-based inspection systems, leading to improved safety, efficiency, and compliance with regulatory standards. 13 2.3 Fire Risk Monitoring and Mitigation Technologies 2.3.1 Current Landscape Experts found numerous emerging fire risk technologies that may apply to Canadian railway rights of way and equipment operations. Because of climate change conditions around the world and particularly in Canada, fire risk research is an evolving area of interest. The current landscape of fire risk monitoring and mitigation for railway operations in Canada relies mainly on individual railroad operating rules and regulations set forth by Transport Canada. Locomotives represent the highest equipment fire risk while railcar braking operations, bearings, and wheels (e.g., slide flats) create additional risk factors. Modern freight locomotives and railcars do not have equipment to detect and automatically report fires in a real-time scenario, therefore they rely primarily on manual inspection or wayside detection (e.g., hot bearing detectors) notifications for mitigation and monitoring of equipment. The development of machine vision applications with camera-based systems used to inspect the various components on railcars and locomotives has advanced, and the systems are being implemented on the railway network in North America. A thermal image acquisition module that provides accurate spatial visualization of locomotive and railcar undercarriage heat propagation is under development. This module can be integrated with existing systems already deployed in North American and may provide additional inspection capabilities related to fire risk monitoring and mitigation. Performance specifications for tank cars carrying hazardous materials can be found in 49 CFR 179 – Hazardous Materials Regulations. Liquefied natural gas (LNG)/compressed natural gas (CNG), propane (LP)/ammonia, and methanol are transported in approved tank car designs (e.g., Department of Transportation [DOT] 113, DOT 112, and DOT 117 designs). Rule 49 CFR 179 currently allows the transport of LNG by rail, but the puncture resistance of the DOT- 113C120W9 design has not been fully established. For passenger equipment, water mist systems using on-board fire detection sensors have been shown to work well in enclosed locations, such as locomotive engine rooms and passenger spaces. It is not currently known through research and testing how a water mist fire suppression system would survive in major accidents or derailments where there is potential for physical damage to the systems, or how the system would perform when fires start on the exterior of the locomotive or passenger railcars. High fire risk track maintenance activities include rail grinding, cutting, and welding. Currently, the assessment of fire risk conditions (e.g., weather conditions, topography) for planning and the use of welder tents, fire prevention, and suppression trains with water and fire- retardants based on application conditions, all provide a vital role in fire risk mitigation. There is not an industry standard for fire risk ground condition susceptibility mapping (e.g., UAVs, satellites) or advanced track materials (e.g., fire retardant coatings) for track right of way and land adjacent to railway track. 14 2.3.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Mapping/imaging technologies using UAVs and satellite to evaluate high risk fire areas on railroad rights of way. Mapping/imaging systems capable of detecting high risk areas without human oversight is a new technology area for railroads. Currently, a platform designed to provide early detection and prevention of forest fires (WildFireSat) is being developed in Canada. Correlation of ground imaging/mapping information with conventional ground condition data for practical decision making can help determine the cohesiveness of current railroad fire risk operating practices with the information these new technologies provide. The proposed research has the potential to provide faster, more detailed geospatial information than current practices on ground conditions to mitigate fire risk. Improved information related to 1) forecasting to proactively identify upcoming risk areas and 2) communicating this information to operating personnel to enable planned actions per operating procedures can be enhanced. Fire retardant coatings for infrastructure, such as bridges and ties in high-risk fires areas. Currently, most railroads have fire retardant products that can be added to water tanks and applied to the right of way infrastructure along with water for fire prevention. There also have been developments in the pretreatment of railroad wood products with fire- retardant intumescent coated mesh. The coating is designed to protect wood in case of fire by forming a protective barrier that stops the spread of flames. Fire retardant products can be further developed and evaluated to determine practical applications for fire risk mitigation. Materials with higher fire risk, such as wood, should be evaluated under real operating scenarios (e.g., distance, ground conditions, and temperature ranges) to determine feasibility. Understanding the life cycle effects of the fire- retardant coating as well as the effects on railroad wood components vs. conventional components. Increased infrastructure safety in high-risk fire areas. Possible reduction in the application of conventional fire- retardant products (e.g., labor, material, mechanized equipment) during high- risk fire seasons. Machine vision camera- based inspection systems (wayside portals) with undercarriage thermal image acquisition modules. Commercial tests have evaluated locomotives instrumented with onboard thermocouples and data loggers to record real-time temperatures for the This system has the potential to deliver early and reliable fault detection for several critical freight and passenger railcars mechanical There are wayside portals implemented on the Canadian railway system that can integrate thermal image undercarriage inspection systems. 15 driveline components of the locomotives. Results showed a good correlation between the undercarriage thermal image module and onboard temperature measurements, suggesting a strong potential for using thermal scanning technology to monitor the major driveline components of locomotives. components, such as rotors, wheels, and traction motors. Improving the accurate spatial visualization of heat propagation (thermal images) of locomotive and railcar undercarriages. The primary benefit of this technology may be the increased inspection cycles of locomotive undercarriages during real-time operations as a preventative fire mitigation measure. Safety and reliability testing of onboard water mist fire suppression sensor systems for locomotives and passenger railcar systems. Research has been focused mainly on the feasibility and applicability of water mist systems to the rail environment and how they can be used in passenger railcars and locomotives. These systems are used in the rail environment on a much larger scale internationally, mainly in European countries. European trainsets are mainly equipped with systems in the passenger compartments. North American systems are designed for the locomotive engine compartment. Safety and reliability testing of water mist fire suppression system designs with onboard sensors. Crash test simulations may provide valuable information into 1) how these systems would survive in major accidents or derailments where there is potential for physical damage to the system, or 2) how the system would perform when fires start outside the locomotive or passenger railcars. Exo, a Canadian based public transit, currently has water mist fire suppression systems that are exclusively locomotive. Generally, when fires propagate to the interior of the locomotive or passenger car, it is possible that a water mist system 1) could have reduced the impact of fire and smoke on passengers (e.g., smoke inhalation and egress time) or 2) could have reduced damage caused by the fire incident. Testing scenarios to determine safety and reliability of the systems may support the implementation (new builds/retrofits) of onboard sensor fire monitoring technology systems. Improved tank car designs to mitigate the potential safety risk of fires occurring from a derailment (transport of LNG by rail). Existing models have been developed with techniques for tank car puncture assessment under various impact conditions to the DOT- 113C120W9. Evaluation of relative performance to other designs can be referenced in report There is a current rule that allows transport of LNG by rail, but the puncture resistance of the DOT-113C120W9 design has not been fully established. Additional testing may provide a better specification (design) to reduce risk of building a Reduce the fire risk of HazMat releases for LNG transport. Prevent potential casualties from derailments. Reduce liability costs for railroads. 16 DOT/FRA/ORD-13/17. Reference/refinement of the existing constitutive model for cryogenic ASTM A240 304. fleet of LNG tank cars with insufficient puncture resistance, requiring changes in future rule making and obsolescence. 17 2.4 Flooding/washout Risk Monitoring and Mitigation Technologies 2.4.1 Current Landscape The SMEs observed multiple research areas related to flooding/washout risk monitoring and mitigation. One research area in-particular that has the potential to provide value to railway planning, with further research, is climatic forecasts on a regional basis across Canada. Current research shows that climatic forecasts are becoming more regional, but additional detailed studies are still required for the research to be useful for any localized analysis. There are multiple monitoring technologies for flooding/washout risk monitoring, such as satellite, UAV, LiDAR, and real-time monitoring. Each of these technologies has its benefits and limitations, but each has shown significant improvements over the past decade and will continue to do so. In general, progress has been based on improved technology and data processing algorithms, but both have limitations of scale and currently have difficulty being applied on a system-wide basis due to installation and material costs and large amounts of data collection and processing. As these technologies progress, justifying their application on a system-wide basis should expand as well. There have been limited numerical studies into flooding/washouts because these studies often require a coupled computational fluid dynamics-discrete element method (CFD-DEM) analysis that is extremely complicated and computationally expensive. Recent advancements in the theory and computational abilities have started to make these analyses more practical, but a stronger fundamental understanding of input properties would be required before these analyses could be become mature. Risk-assessment studies have been performed (e.g., Amtrak), but there is a lack of quantifiable values on the force side (flood elevation, water force) and strength side (washout resistance) that may make decisions on remediation difficult. 2.4.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders The development of regional climatic models that forecast changes in heat/cold/precipitation in different areas of Canada. Amtrak and Canadian Pacific Kansas City (CPKC) have looked at regional climatic models (Amtrak) or had consultants perform their own (CPKC). The data collected from these models could be used for culvert/drainage design and risk-assessment and the identification of changing risk regions. Optimal culvert designs could be established, minimizing the risk of track washouts. Using forecasted regional data for flood and washout risk assessments/ mitigations can increase the safety of railway operations and focus efforts where needed. Continue researching how satellite (e.g., Interferometric Synthetic Aperture A small range of studies have been conducted using a combination of satellite Data collected from the technologies can be used to inform flood/washout risk Optimal inspection methods will lead to enhanced safety, minimized 18 Radar [INSAR]), UAV, and LiDAR technology can be used to address a range of flooding/washout concerns. This may include development of algorithms or techniques. images, UAV, and LiDAR technology for flood monitoring and water inspection. management and mitigation with benefits such as providing real- time information. infrastructure damage, and more efficient operations. Physical monitoring devices that provide real-time alerts are already available. These technologies may or may not need additional documents and support. Some research has been done on water level sensors and flood monitors to enable remote monitoring. Improved alert systems and use of alert systems in a system. Improve safety of railroad network. Modeling the washout process is difficult and still requires fundamental research and testing. Modeling includes categorizing 1) the various “washout” mechanisms (what causes the washout) and 2) the important parameters that influence washouts, and quantifiable numbers to represent forces and track strength, including general values or a detailed list depending on the embankment material and strength properties. Some research has been conducted on modeling the washout process, including a model for railway ballast washout and idealization of ballast for dynamic characteristics. Understanding the washout process and how it impacts railway infrastructure will allow for mitigation measures to be established and put in place. This research can inform the safe design of railway track and embankment to reduce the impact of washouts. For example, quantify how various remediations would improve embankment strength and reduce risk of washouts. Data from the models can inform industry of best practices for washout mitigation and management. Development of tools or framework to assess flooding/washout risk. There has been a BNSF pilot study on washout risk. The goal of this tool or framework would give guidance regarding the important parameters for washouts and allow railroads to rank locations. The most at-risk locations would be identified and prioritized for increased safety measures. 19 2.5 Geohazard Monitoring/Mitigation Technologies 2.5.1 Current Landscape Using geohazard monitoring/mitigation technologies for landslide prediction is extremely difficult, especially in remote mountainous areas. There are a range of technologies and techniques that can monitor/detect landslides (e.g., InSAR, UAV, global navigation satellite system-real-time kinetic positioning [GNSS-RTK], borehole inclinometers/piezometers). Each technique has its own strengths and limitations, and dataset combinations have often been used in past research projects. For large networks, satellite InSAR appears to be the most promising method because it allows monitoring of locations with consistent satellite coverage without a human site visit. The number of locations is growing with the ever-increasing number of satellites, but it remains difficult to interpret InSAR movements and quantify large datasets using only an overhead view. In addition, InSAR will likely be one of many methods used in railroad landslide monitoring programs. The UAVs have the benefit of viewing a terrain from multiple angles and the ability to set specific flight paths, but this method requires a site visit, and it may be limited by weather conditions and regulations. The GNNS-RTK technologies are able to continuously monitor sites and capture physical slope movements, but they are limited by only having a single location per instrument and difficulties uploading or collecting data. All the listed technologies are anticipated to become more advanced in their capabilities in the next decade. Due to various limitations of each single technology, combining datasets provides the most accurate assessment but doing so requires an amount of effort likely reserved for only a few high-risk locations. Satellite technology may be used for a large-scale desk study that identifies a manageable number of more specific locations for site visits and additional instrumentation. The Canadian Railway Ground Hazard Research Program (RGHRP) appears to be a world leader in the development and use-cases of these technologies, including case studies in the Assiniboine River Valley and Thompson River. Other measurements that may have certain applications but are less matured for geohazard monitoring are fiber optics and other sensor systems that are often already installed in railway track for other purposes. For rockfalls detection/prediction, LiDAR, UAVs, and other technologies have shown significant advancement in the past decade. These methods are often not currently able to be implemented at-scale, and there are parallel efforts to develop frameworks that can identify high-risk locations across a subdivision/network. 2.5.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Support and develop algorithms that can monitor/detect/predict geohazards. This may require additional site locations and datasets. Research has compared the performances of swarm-based algorithms such as Particle Swarm Optimization (PSO), Ant Colony Development of an algorithm that can process data and serve as an early detection system for geohazards. The development of sophisticated algorithms can provide rail operators with valuable data and insights, enabling more informed decision- making regarding route 20 Optimization (ACO), and Genetic Algorithm (GA)-optimized adaptive neuro-fuzzy inference systems (ANFIS) in the spatial prediction of landslides. These methodologies can be adapted and applied to the rail industry for geohazard detection and prediction. planning, infrastructure investment, and risk management. Support and continue to advance technological advancements in the wide range of monitoring technologies (InSAR, UAVs, GNSS- RTK). Studies have used remote sensing technologies, such as InSAR, LiDAR, and UAVs for monitoring terrain deformation, identifying potential landslide areas, and assessing the extent of damage after a geohazard event. Specify the type of technology that will be used, identify the geohazard of interest, and come up with an integration of said technology in a geohazard monitoring application. Improved geohazard monitoring can significantly enhance the safety of rail operations by providing early warnings of potential hazards, reducing the risk of derailments, and ensuring the safety of passengers and freight. Development of tools or framework to assess geohazard risk. The goal of this tool or framework would be to give guidance for the important parameters for geohazards and allow railroads to rank locations to help identify the most at-risk locations. Development and operation of geohazard management programs focused on risk-based rock fall rating systems for railway operators, such as the program developed by BGC Engineering for CN Rail Establish a mapped, risk-level based database of railroad locations that can serve as guidance for railroads to determine navigation procedures or risk mitigation measures. The tools or frameworks will provide rail operators and stakeholders with valuable data and insights to make informed decisions regarding infrastructure investment, maintenance, and risk management. 21 2.6 Technologies for Managing Cold Weather Operational Challenges 2.6.1 Current Landscape Due to significant safety and operational challenges, winter conditions and cold temperatures can negatively affect the performance of both railway infrastructure and rolling stock. Managing winter operational conditions is vital to maintaining a fluid and safe railway network. Recent technological developments show great potential in helping railway operators overcome these challenges. The technology scan references for this theme can be grouped into three general sub- themes: 1) railway infrastructure under winter conditions, 2) passenger rail operations in winter environments, and 3) brake system performance and safety for freight trains. Research has shown success in long-term monitoring of continuous-welded rail (CWR) by measuring lateral displacements of the track in curves using a sensor type that resists low temperatures and measurement drifts. Field monitoring of frost heave formation and numerical modeling was carried out to observe the effect of culverts on frost zones, frost heave and thaw softening, and their impact of culverts on the wheel/rail interface. The studies, which have been concluded, were contained to a single site under well-maintained track. It is unknown how the method would perform for lower-class tracks, indicating the need for further study. Compared to real-time monitoring of the track substructure, vehicle-mounted ground penetrating radar (GPR), and the ability to conduct mobile measurements across the railway network, has proven cost effective in the identification of frost heave sources. Mechanisms of frost penetration under railway embankments, protective snow fence strategies, and a literature review on icing effects on railways provide insights for winter mitigations. A thawing-guided drainage system to prevent icicles in old railway tunnels was developed and fabricated in the lab. Testing showed this system was effective against icicle buildup, but no plan was mentioned to make the system available commercially. A review of construction techniques summarized modern methods for reducing the environmental impact of railway construction on permafrost. Studies on thermal- moisture-mechanical behaviors and thermoelectric heater systems offered solutions for mitigating icing on track components. Researchers developed rolling stock designs to reduce winter icing and ballast splashing. While studying brake systems, researchers found that the braking distance increased with ice buildup on the brake shoe and abnormal disc wear could be reduced by optimizing train driving strategy. Most of this research was not conducted in Canada, raising questions about its applicability to Canadian winters. There are information gaps in understanding 1) the effect of cold temperatures on non-contact ultrasonic rail flaw testing, 2) the root cause of icing on railway infrastructure, 3) the autonomous detection and effective mitigation of compressed air leaks on trains. An emerging technology to prevent undesired train movements, such as the automatic parking brake, has yet to be thoroughly tested by the railway industry. Further research is needed to address brake system performance degradation, especially in extreme cold temperatures. 22 2.6.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Brake system performance and safety for freight trains Acoustic brake system leakage detection should be validated on a larger variety of equipment types. The detection system should also be tested in uncontrolled field environments to assess its feasibility for wider adoption. A proof-of-concept system to autonomously detect compressed air leaks on moving trains was developed and a small amount of data was collected. System validation and further testing in uncontrolled field environments could lead to wider adoption. A better understanding of and detection method for air brake leakage could lead to safer, more efficient locomotive operations. Other leakage detection methods should be investigated. A literature review should be conducted to summarize the current best practices from other industries that rely on pressurized pneumatic systems. A proof-of-concept acoustic brake system leakage detection method was studied. Understanding the best practices and identifying the present gaps in information could lead to more targeted studies and tests. More advanced leakage detection methods can increase safety in locomotive operations. Research and tests on methods to prevent undesired movement from brake cylinder leakage in cold temperatures are needed. Potential devices should be tested both in the laboratory and on trains. Transport Canada is conducting cold weather air brake research in collaboration with NRC. Data regarding the performance of air brakes in temperatures ranging from 0°C to - 40°C has also been collected. Help industry and government determine a common performance specification and understand potential operational issues. Could lead to regulations or best practices that will increase the safety of rail operations in cold weather. Railway infrastructure under winter conditions Assessment of measurement techniques, such as comparing the performance of different sensor types by measuring the same curves, to monitor lateral track displacement in sharp curves should be conducted in the Canadian environment. The data can be Previous research has been done on the remote monitoring of continuous welded rail on curves in cold temperatures. An online monitoring system and mathematical model of ultrasonic stress detection was studied. Optimal measurement techniques can be determined for measuring track displacement in sharp curves due to cold weather. Understanding the effect of cold weather on track displacement can help the industry determine at risk areas and ensure safe locomotive operations. 23 compared against track geometry car measurements to see if noticeable curvature changes are recorded. Existing data from curve measurement and frost heave monitoring research should be analyzed to find trends and early warnings, potentially with the help of advanced statistical methods such as machine learning. Frost heave monitoring techniques developed outside Canada should be validated in Canada. Research conducted in Canada at the VIA Rail test site for long-term frost heave monitoring should be expanded to validate their findings and applicability in various locations and under different axle- load environments. A better understanding of the frost development mechanism and its impact on safety and performance of train operations. Industry and government will better understand the effects of frost heave on track geometry, allowing mitigation measures to be determined and safety to be increased. Further finite element model development is needed to understand better the root cause of tie plate icing and its mitigation. The thermoelastic heating system could be tested as one potential solution to this problem. An investigation into tie plate/ice jacking was conducted by Transport Canada and MxV Rail. This could result in a better understanding of the mechanisms causing tie plate icing and the use of modeling to provide the means to predict the occurrence of tie plate icing. Understanding the effects of tie plate icing can help guide the industry’s winter preparedness and remediation methods, both of which will lead to safer operations. The effect of cold temperatures on non- contact ultrasonic rail flaw detection accuracy needs to be better understood. One way this can be achieved is by conducting measurements on a much larger rail sample in the field and in the environmental chamber. The feasibility of methods to mitigate the effects of cold temperatures on ultrasonic sensors could also be studied. A study was conducted by Transport Canada and MxV Rail on ultrasonic rail flaw parameters in extreme cold temperatures. Better understanding of temperature effects on ultrasonic rail flaw detection could result in a better calibration method of ultrasonic rail flaw detection in the winter that results in more accurate detections. Railway infrastructure managers will be able to detect rail flaws in cold temperatures with higher accuracy, leading to more targeted remediation and improve safety. Methodologies to reduce snow and ice buildup along railway infrastructure can be investigated. A starting point could be adapting Research includes an assessment on old railway tunnel icicle prevention, de-icing catenary wires, and an investigation on snow Developed methods and mitigation measures to reduce snow and ice accumulation along railway infrastructure. Increased safety and efficiency of locomotive operations. 24 foreign research methodologies into the Canadian winter environment and assessing their merits. accumulation on bogies. Passenger rail operations in winter environments Snow and icing shields could be installed on Canadian passenger rolling stock to compare their effectiveness against vehicles that did not have these installations. There is research on anti-snow structures, such as deflectors and wheelset snow shields. Knowledge of the effectiveness of snow and icing shields and how they impact rolling stock operations. Could lead to more efficient rail operations. Stopping distance tests and long-term brake disc monitoring due to abnormal wear in winter conditions should be conducted in Canada. Some research has been performed on abrasive wear of brake discs. An understanding of the effect of the staged driving strategy on safety and methods to reduce abnormal brake disc wear. Addressing abnormal abrasive wear due to winter conditions will lead to more consistent braking and train control in winter conditions. Novel catenary and trolley wire deicing systems proposed in the research reviewed should be installed in Canadian environments to validate their effectiveness. Previous research has been based on the use of heaters. This method provides an alternative method of deicing trolley wire. This method will lead to a decrease in energy consumption of deicing and contribute to industry decarbonization efforts. 25 2.7 Technologies for Subgrade Stabilization 2.7.1 Current Landscape From geotextiles (e.g., geogrid, geocells, various geotextiles, such as Tracktex or Mirafi) to drainage (e.g., trench drains), a wide range of technologies provided by multiple manufacturers, distributors, and consultants are available for subgrade stabilization (e.g., grout, polyurethane, geospikes, rail piles). Each technology comes with benefits and drawbacks (e.g., improve subgrade strength versus add confinement versus barrier layer), and different technologies may be suitable in different situations depending on the issue root cause, the issue severity, the allowable track time, the track access, and the cost, among other considerations. While there is a range of available products and most of these products are mature technologically, there are gaps in guidance regarding 1) which product would be appropriate for a particular situation and 2) the effectiveness and longevity of each product. Most of this information is provided by the vendors themselves, and it can be difficult for railroad agencies to compare different technologies and have reliable data to make decisions. 2.7.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders A scan of the information from various railroad agencies (freight, passenger, transit) to identify common subgrade issues the railroads currently experience to understand the current technologies those agencies employ along with specific needs. Limited existing research. Guidance for more targeted research and insight into specific hurdles that must be crossed to make promising technologies implementable. Improved documentation and knowledge of technologies can better justify use and reduce subgrade issues. Guidance on industry- accepted identification of subgrade issues, best practices, and tradeoffs of different subgrade technologies. MxV Rail has an internal document summarizing subgrade issues and various remediations. Much other research focuses on one or two technologies. Improved railroad decision making regarding subgrade. Reduced risk of subgrade issues Testing, instrumenting, and documenting different subgrade stabilization in a track environment. MxV Rail summarized a 30-year long test looking at different applications to stabilize weak subgrades (geocell, Hot-Mixed Asphalt) and worked with Class 1 railroads Improved railroad decision making regarding subgrade Reduced risk of subgrade issues 26 on two specific issues (ballast pocket and flooded cut) and monitored the benefits of the remediations (geogrid and various barrier geotextiles). MxV Rail is currently monitoring five technologies to reduce mud spots with another Class 1 railroad (unpublished). There is a general lack of third- party documented research on performance. 27 2.8 Enhanced Train Control 2.8.1 Current Landscape Enhanced Train Control (ETC) systems are being developed to enhance the capabilities of the train operators, to reduce human error, and to improve the safety of both passenger and freight railway operations in Canada. The ETC technologies can alert the train crew to danger and, at their highest functionality, slow or stop a train to prevent a collision or derailment. The ETC will be a communications-based train control (CBTC) system, critically dependent on dependable connectivity among fixed and mobile segments. Otherwise, when operating in the highest functionality mode, ETC can stop trains unnecessarily. Redundant diverse communication media are required to achieve the necessary level of dependability. Communications must also employ cybersecurity to avoid compromising safety. The minimum required CBTC functionality varies from country-to-country due to differences in objectives, needs, and constraints. The Interoperable Train Control (ITC) Positive Train Control (PTC) system can be leveraged because it is the most widely deployed CBTC system in the United States (U.S.), and it is already installed on many Canadian locomotives that operate across the border since they must be interoperable. Advanced CBTC modes such as Enhanced Overlay-PTC (EO-PTC), Quasi-Moving Block (QMB), Full Moving Block (FMB), and Automated Train Operation (ATO) add functionality that can significantly improve both the railway throughput and the average velocity along with increasing safety. The National Transportation Safety Board (NTSB) and the Federal Railroad Administration (FRA) have identified certain classes of human error that PTC (and therefore, potentially ETC) could mitigate, including: • Preventing train-to-train collisions during restricted speed operations. • Automatically returning PTC to the active mode following switching operations (or alternative means of collision protection during switching operations). • Eliminating the possibility of human error when identifying the track occupied by a train (e.g., during initialization) in multiple track territory. • Eliminating the risk of miscommunication between dispatchers and roadway workers in charge when establishing working limits and PTC protection. CBTC is a complex distributed real-time safety-critical system requiring sophisticated testing. The technology scan references and opportunities for future research fall into four general sub-themes: 1) existing solutions that can be leveraged, 2) critical communications issues, 3) required or minimum essential functionality, and 4) unique testing needs. 28 2.8.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Existing solutions that can be leveraged. The needs, requirements, and constraints for ETC should be compared with ITC PTC specifications and Interface Control Documents (ICDs) to identify commonalities and prioritize gaps. Based on the gap analysis, potential solutions should be proposed and assessed. Experience and specifications resulting from historical use of ITC PTC on multiple U.S. freight railroads. ETC requirement specifications that are optimized toward the specific needs of Canadian railroads. Maximization of ETC return on investment (ROI) by minimizing the amount of new developments and safety certifications required while meeting critical unique needs. Existing solutions that can be leveraged. The potential cost vs. benefits of energy management systems (EMS) should be evaluated in the context of Canadian railroad applications. EMS assists drivers or directly controls throttle and brake to save fuel and to manage in-train forces. Performance results from historical use of EMS with ITC PTC on multiple U.S. freight railroads. Interoperability and performance standards for EMS systems and their integration with ETC. Maximization of ETC/EMS ROI by minimizing the amount of new development required while meeting unique needs. Increasing the fuel- efficiency of train operations across multiple railroads. Reducing equipment wear and damage by reducing in-train forces. Required or minimum essential functionality. The potential cost vs. benefits of integrating a dispatch movement planner/scheduler with EMS pacing should be evaluated in the context of Canadian railroad applications. Historical performance results from historical use of a movement planner together with PTC (e.g., from NS Railroad’s experience). Assessment of ROI from integrating a dispatch movement planner/scheduler with EMS pacing. Optimal train movement planning integrated with EMS and real-time train location reporting can improve fuel efficiency and reduce rail network congestion. Required or minimum essential functionality. Potential advanced train control modes that can increase the ROI and the safety of ETC should be assessed when specifying the Reports and specifications from Higher Reliability and Capacity Train Control (HRCTC) and ATO projects. Estimate of ROI from incorporating ETC into each of the following advanced operating modes: EO-PTC, QMB, FMB, and ATO. Solutions that can increase railroad safety, increase capacity, increase network velocity, increase productivity, reduce headways, and reduce delays. 29 functional requirements for ETC. Required or minimum essential functionality. The following safety enhancements should be assessed when specifying requirements for ETC: • Collision prevention at restricted speed • Non-overridable enforcement of stop at entrance and modifiable speed restrictions for work zones • Automated track discrimination in all multi-track scenarios. Reports, specifications, and test results (in some cases) on QMB, FMB, Employee-in- Charge Portable- Remote-Terminal (EIC PRT), and Positive Train Location (PTL). Assessment of ROI from providing the cited safety enhancements to ETC. The potential enhancements cited, if incorporated into ETC, can protect against additional forms of human error, thereby increasing railroad safety. Critical communications issues. Evaluate/compare private vs. commercial networks, open vs. proprietary waveforms / protocols, dedicated vs. shared vs. unlicensed radio frequency (RF) spectrum, and terrestrial vs. satellite in terms of performance, cost, and interoperability among railroads (domestic and below the border). Historical performance of ITC PTC radio communications systems used in the U.S. (dedicated PTC radio, WiFi and cellular) as well as mobile communications systems available in Canada. Assessment of the adequacy of existing coverage and infrastructure (towers, satcom, power source, backbone) along tracks in Canada. Identification of what is most feasible to add – means to operate where coverage (terrestrial) or visibility (satcom) is low. Determination of the most cost-effective solution to achieve the required highly reliable and interoperable communications on Canadian railroads. Critical communications issues. Survey RF propagation prediction and message traffic load models to identify best suited for ETC. Propagation models used by cellular companies have benefited from far more development funding than railroads can provide. A message traffic load model has been developed specifically for ITC PTC and applied (by MxV Rail) to the radio network design in all large U.S. urban areas based on historical and predicted train operations. These Determine the best suited existing propagation model(s) to predict required quantity, locations, effective radiated power (ERP), and frequency of fixed ETC radio sites. Determine whether any modifications to the message traffic load model are necessary to support ETC. Avoidance of cost to develop new propagation and message traffic load models for ETC. 30 models have been used to achieve successful deployment of ITC PTC. Critical communications issues. Trade off train location determination system (LDS) solutions regarding Canadian railroad needs. Historical performance and costs of GNSS- based LDS (ITC PTC) vs. Balise-based LDS (European Train Control System [ETCS]) Assessment of performance and costs (ROI) of GNSS vs. Balise-based train positioning and identification of associated options for ETC. Selection of solution that offers low life cycle cost while meeting critical system performance requirements. Critical communications issues. Assess potential cybersecurity threats to ETC comms and GNSS for the present and foreseeable future. Determine the level and type(s) of solutions needed to protect ETC against the potential threats (e.g., authentication, encryption, key management, spatial discrimination [via antennas]). Information (e.g., from actual incidents in more than just railroads) on the increasingly sophisticated potential jamming/denial of service and spoofing threats. Historical performance of various candidate threat mitigations. Available GNSS vulnerability test bed (at MxV Rail). Design of functions and equipment to detect, log, analyze & mitigate the threats (real time vs. post). Identify which messages require encryption vs. those only needing authentication. Test results under various jamming and spoofing scenarios to optimize key parameters. Cost-effective cybersecurity threats mitigations. 31 2.9 Technologies for Increasing Grade-Crossing Safety 2.9.1 Current Landscape Numerous studies on technology used to detect and monitor grade crossing obstructions (e.g., trespassers, crossing violations, stalled vehicles, etc.) have been and are being conducted to improve safety and reduce incidents. These studies are primarily focused on the ability to detect instances where a person or object is 1) within the limits of a rail crossing and 2) at risk of being struck by rail traffic (freight, passenger, or transit). The studies have primarily been proof-of- concept in nature to determine whether machine-vision based equipment is a viable option for collision prevention at grade crossings. Information gaps exist regarding connecting these devices to current safety warning and/or dispatcher/train systems. There is limited practical application for these technologies due to train stopping distances, the inability to quickly remove stalled or high-centered vehicles from crossings, and the potential for false alarms. Fully deployed detection/warning systems are not currently in use on a large scale. Also, due to costs and lack of existing infrastructure, many of the technologies available and currently being researched have power and communication requirements that make application in remote areas potentially difficult. While the detection of crossing violations and near-miss events has been achieved, the prevention of crossing violations altogether remains unresolved. While detecting obstructions and crossing violations is a step toward achieving greater industry safety, the prevention of crossing violations and obstructions/trespassing can and should be a topic of continued investigation. In Europe, the use of connected vehicle technology is being deployed to provide in-vehicle warnings and/or prevent a motor vehicle from entering the crossing zone. This technology takes a different approach to the human factor problem or eliminates the human factor completely. In the U.S., for at least three years now, some railroads have been working in conjunction with the WAZE navigation application to provide users of the app with warnings near railroad grade crossings. A couple railroads working in this area include Metro-North and Southeastern Pennsylvania Transportation Authority (SEPTA) (commuter lines in the New York City and Philadelphia areas, respectively.) In terms of improving grade crossing safety, grade separations and crossing closures, as well as sight-distance improvements, also need to be considered. Both a cost-benefit analysis of these improvements and traditional warning systems are covered in the current Federal Highway Administration (FHWA)/FRA methodology (GradeDec) and an ongoing National Cooperative Highway Research Program (NCHRP) project to update that methodology. 2.9.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Continued investigation and development of detection systems with a focus on deployment of fully functional Extensive preliminary research on machine vision, cameras, LiDAR, etc., to detect Better detection algorithms and technologies for vehicles stuck or stalled on crossings, possibly Reduce grade crossing incidents involving stuck or stalled vehicles, possibly pedestrians. 32 systems in areas with known problem crossings. objects in crossing zones. also for pedestrians lingering at crossings. Development of system interfaces for warning systems needs to be continued and integrated into existing railroad warning and alert systems already in use. Tends to be supplier driven. Limits on how many accidents can be prevented due to train stopping distances, etc. Better detection systems for vehicles stuck or stalled on crossings, possibly also for pedestrians lingering at crossings. Reduce grade crossing incidents involving stuck or stalled vehicles Expanded research into the human factor side of crossing violations is needed. Much work done by FRA and more work ongoing. Better understanding of factors likely to improve grade crossing safety. Reduce grade crossing incidents involving drivers likely to ignore existing warnings or take extreme risks Cost/benefit analysis of different violation- deterrence technologies to enable railroads or transit agencies to select the most cost-effective grade crossing enhancements for a crossing based on potential benefit gain. There is a current NCHRP project focused on updating the factors used for such analysis. Methods for more efficient allocation of limited resources for improving grade crossing safety. More efficient allocation of limited resources for improving grade crossing safety. Development of a modularized detection system that would not only detect potential crossing safety violations but prevent collisions or accidents from occurring once detected and that could be integrated into existing railroad safety systems. Not aware of any previous work. See implementation challenges discussed in current landscape. Limited value proposition at this time due to challenges discussed in current landscape. 33 2.10 Technologies for Monitoring Trespassing 2.10.1 Current Landscape Experts grouped technologies into three general sub-themes; 1) video-based technologies for detection, 2) other technologies for detection, and 3) the broader overall topic of trespasser prevention. The video-based trespasser detection systems are primarily wayside-based and employ machine vision systems with neural networks or other AI algorithms. One variation on the video-based system is the use of a thermal vision camera, which can distinguish features that might be hard to detect with conventional video. Wayside vision-based detection systems should be placed at known hot-spot locations to be used most effectively. A UAV-based system might help identify trespasser hot spots, but the system currently employs manual detection. One video system noted the potential for an on-board application to help identify hot-spot locations. Compared to wayside systems, the constantly changing background presents a challenge to in-motion and on-board video systems due to the additional effort required to indicate trespassers as well as to distinguish trespassers from right-of-way workers. Other detection technologies noted were fiber-optic acoustic sensing and radar. Both technologies seem to require further development compared to the video-based systems. A fiber- optic system is inherently wayside-based only. All the monitoring and detection technologies share the challenge of making effective use of the findings. An on-board system might help locate trespasser hot spots, although it is likely that railway train operators and maintenance-of-way personnel can already provide such information. Once a wayside system detects a trespasser, either a warning (visual and/or audible) needs to be initiated (especially if a train is in the vicinity) or enforcement needs to be called. In some cases, it might be more cost-effective to install an automated pedestrian crossing. References covering the broader theme of trespasser prevention provide further information regarding the many other factors and challenges to the implementation of various mitigation strategies, technological or otherwise, that need to be considered. Trespassers, even those with cell phones, are not likely to be using a navigation app such as WAZE that can provide highway- rail grade crossing warnings, and apps such as WAZE or other technology-based detection and/or warning systems are likely to have little or no effect on trespasser fatalities, a significant portion of which are suicides. Recent research indicates the number of suicides as likely to be undercounted, and educational efforts to deter trespassing and raise awareness of the potential dangers of railroad crossings could have the undesired effect of attracting more suicides to railroad tracks. While the various efforts in recent decades have reduced grade crossing incidents and fatalities significantly, challenges to reducing the number of trespassing issues on the tracks remain. Further investigation into other methods and technologies for trespass prevention should include trespassing deterrence methods, such as thick/thorny vegetation, fencing, and pedestrian underpasses and overpasses. Each method has its respective advantages, challenges, and costs. Urban planning can also play a part by minimizing the placement of popular pedestrian destinations, such as convenience stores, across the tracks from common pedestrian origination sources, such as apartment complexes. 34 2.10.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Further development of video- based technologies, including on-board and UAV-based technologies is required, in addition to fixed wayside-based systems. Several studies currently underway. Better detection technologies for trespassers on railroad right-of-way Better response to trespasser incursion on railroad right-of- way (pending implementation as noted below) Next-stage development/implementation of detection technologies to coordinate with enforcement, initiate warnings, or take other appropriate action. Not aware of any but might be happening with suppliers. Proposed methods for implementing trespasser detection with existing protocols and systems. Better response to trespasser incursion on railroad right-of- way. Continue, reinvigorate, and/or update the Community, Analysis, Response, and Evaluation (CARE) model effort has previously undertaken. Some of the limitations and challenges noted in the previous effort might now be able to be addressed with some of the new tools and technologies available. Past work on CARE model by TC and FRA. Efforts will likely include solutions in the enforcement or education spaces, not just engineering. Reduce frequency of trespassing incidents. 35 2.11 Development of Tools and Analytics for Risk Assessment 2.11.1 Current Landscape The SMEs reviewed available research references related to quantitative risk assessment (QRA) for hazardous material transport. The references suggest that, in the transportation industry, QRA is primarily used for public safety and emergency response preparedness. Risk assessment can be grouped into security, emergency response, public safety, and operating safety. The rise in large datasets produced by centralized traffic control, wayside and onboard detection systems increase the available factors that contribute to a QRA model. Other datasets, including transportation routing information, rail incident information, citizen population densities, environmental data (including environmentally sensitive areas or geohazards), grade crossings and usage, and general orders of operation for railroads can provide useful additional information to identify critical factors in a risk assessment. An example of the use of QRA related to rail operations and safety that used a section of rail in the Canadian Cordillera Mountain range was referenced. The assessment in the example used Monte Carlo simulations to develop a risk probability distribution to estimate the risk to the life of freight train crews operating in this area. Currently, QRAs are being performed on several platforms with a variety of statistical tools. The tools identified by the probability of promising results include Bayesian Networks and Monte Carlo simulations and supplemental tools like computational fluid dynamics (CFD) models, F/N (frequency of event vs number impacted often as derailments or loss of life) curves, and Event Trees for detailed analysis. These tools are available in Python and R, programming languages commonly used in neural networks, machine learning, and statistical analysis. Other analytic tools are available but not referenced in the technical literature. 2.11.2 Opportunities for Future Research Three common areas identified in the current landscape section include the following research opportunities: Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders The need for more detailed transport accident information. QRA relies heavily on accident and incident data of previous occurrences. The available databases of accidents and incidents are often limited in detailed information, leading to a lack of clarity on the cause of incidents. The incorporation of more detailed information in accident and incident data could strengthen QRA and help to better understand causes of incidents and accidents. It could also help strengthen predictive modeling. More accurate QRAs leading to safe and more efficient transportation of goods by rail. Volume of road traffic at grade crossings (intersecting roadway Annual average daily traffic data (e.g., the total volume of vehicle The incorporation of better traffic data at grade crossings could Grade crossings are areas at highest risk of incidents and 36 systems) could provide improvements to the risk models. traffic of a highway or road for a year divided by 365 days is widely available. This data could be incorporated in QRA to quantify the risk of grade crossings on the movement of goods. strengthen QRA and help strengthen predictive modeling. accidents, and accurately quantifying the risk of transporting goods through grade crossings could improve public safety and ensure the efficiency of rail and road transportation by reducing high risk incidents. The domino effect of breached freight cars that are modeled with various neural networks could be included into a QRA model for the potential of added risk that could increase the severity of the QRA model results. There exists a body of research that models the severity of derailments by looking at the potential for damage to cars and breaches that could occur in a loss of containment. Incorporating tank car damage modeling into QRA could more accurately predict the severity of an incident and the appropriate preventative or mitigative measures. More accurately understanding the risk severity of transporting goods could help increase public safety. Written guidelines for input factors that feed the QRA models could aid in producing a repeatable, more accurate, and less biased result. As new information is discovered, the guidelines should be updated to accommodate the new information. Guidelines for QRA exist and are valuable to support the implementation of assessment methodologies (e.g., the “purple book”). Railways could leverage this guidance to help inform their own QRA efforts. Incorporation of best practices from QRA guidelines could strengthen QRA for rail transportation. More accurately understanding the potential risk of transporting goods. Adding other datasets can be added to most QRA models to determine if additional factors may be identified, i.e., localized weather events, peak and off-peak traffic density changes for citizen movements. Review troubleshooting guides for weaknesses in instructions for multiple related or non-related failure events. Work to date on QRA in rail transportation has highlighted a lack of data, and an opportunity to input additional data sets to better understand exposure, consequences, and preventative and mitigation measures. The proposed outcome would increase the comprehensiveness of QRA results and include impacts on additional factors such as sensitive environments and vulnerable communities. A more comprehensive QRA could result in better public safety. 37 2.12 Cybersecurity 2.12.1 Current Landscape The rise in large volumes of data generated by various sources, including, but not limited to, wayside and onboarding inspection systems, signaling systems, and train control systems, as well as the integration of communication-based connected railroad technologies known as the Rail Internet of Things (RioT), has provided significant value to the railroad industry. However, this advancement concurrently poses challenges related to data storage, privacy, data fusion, and cyberattacks. Three main areas were identified in the current landscape of this review: • Rail cyber-physical systems threats and vulnerabilities. • North American railroad industry efforts in cybersecurity. • Cybersecurity measures and mitigation strategies. The reviewed literature delineates various rail cyber-physical systems and associated threats. The systems encompass the areas of train control, traditional railway signaling, Balise-based data transmission, railway traction power, voltage control, human-machine interface, public address or public information display screens, railway sidetrack, ATO, and CBTC. Among the identified threats are: • Multiple attacks during electromagnetic interference. • Denial of service (DoS). • Unauthorized network access. • Message modification. • Man-in-the-middle (MITM) attacks potentially resulting in false data injection, spear- phishing emails, etc. • Attempts at fraud through corporate identity misuse. • Scans for information on corporate executives. • Occasional high-volume or otherwise suspicious activity from foreign internet protocol addresses. • Compromise of email accounts belonging to shippers or other industry entities. • Creation of falsified websites as a cyber-criminal activity designed to lure unsuspecting individuals into providing personal and financial information. 2.12.1.1 North American railroad industry efforts in cyber security The AAR has established the Rail Information Security Committee (RISC) to facilitate the exchange of practices, threats, vulnerabilities, and incident response strategies related to industry-wide information sharing. These strategies include collaboration through the AAR Railway Alert Network, sharing insights with government departments and agencies in select 38 countries responsible for cybersecurity, and benchmarking security against well-established and proven cybersecurity standards. The report focuses on the role of suppliers and their influence in the cybersecurity of rail systems. The outlined practices provided to suppliers cover areas such as software and services, access control, account management, session management, authentication/password policy and management, logging and auditing, communication restrictions, malware detection and protection, heartbeat signals, and ensuring reliability and adherence to standards. 2.12.1.2 Cybersecurity measures and mitigation strategies In addressing cybersecurity measures and mitigation plans, the reviewed literature outlines a range of strategies, including intrusion detection/prevention systems, firewalls, host intrusion prevention systems (HIPS), authentication, secure file transfer protocol, log collection, encryption, dedicated equipment, and the implementation of a physical “air-gap,” among other approaches. The description of supplier life-cycle security programs is designed to establish a framework for developing products with fewer weaknesses and vulnerabilities. These programs focus on identifying and remediating potential weaknesses and vulnerabilities before the installation of software and systems in the customer’s environment. The literature also explores cyber risk management methodologies for RIoT, providing guidance to stakeholders and encompassing key components, such as identifying threat sources, the technical decomposition of architecture and components, prevention, and detection, as well as mitigation and recovery strategies. Lastly, the research presents a model-based cybersecurity analysis, integrating inputs from the security analysis on system components, potential attacks, and attacker models into a comprehensive goal, system, and attacker graph. The findings show that deploying firewalls is highlighted as an effective way to protect critical local area networks (LANs) from compromise. 2.12.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders Resilience and disaster recovery planning. Collaborative initiatives in North American railroads as part of the RISC, as well studies in risk assessment and vulnerability analysis. Opportunities for North American railroads to evaluate the security profile and take action as needed. Railroads can expect to encounter robust cyber threats, so there is a need to evolve security programs given the dynamic risk environment and the potential for increasing the use of technology for safe and more efficient operations. The research also suggests that the railroad freight industry needs to use today’s technologies as the Overview of rail technologies and how these developments had consideration in cyber threats, risk Opportunities exist for North American railroads to improve their cybersecurity practices by This can support the early identification of threats and vulnerabilities as well as the design and 39 foundation for more innovation to further enhance the safety and security of the network. assessments and their components, and communication nodes and links. collaborating with vendors during the design phase of railway technology development. This collaborative approach can enable the integration of cybersecurity measures into the design process, ensuring that security considerations are prioritized from the outset. implementation of prevention and mitigation strategies for cyber threats. Machine learning for threat detection involves exploring the application of machine learning algorithms for anomaly detection, e.g., anomalies in signaling systems, advanced train control systems (ATCS), wayside and onboard detection systems, and positive train control (PTC), among others. This is an emerging research area with very limited information available. Some initial efforts have focused on identifying system vulnerabilities to unexpected disruptions through the application of deep learning techniques. Opportunities for North American railroads to expand the research and applications of using machine learning methods for threat detection. With large amounts of data collected from multiple technologies in the railroad, using machine learning approaches can help support the development of models to identify threats. Conduct comparative analyses of the different model results with independent studies from previous risk assessments. Literature in this area is very limited at the time of the review, and the studies focused more on the framework design without a formal comparison with other approaches. Opportunities for North American railroads to identify the strength and weaknesses of different models so researchers and practitioners can better understand which cybersecurity aspects are more effectively addressed by each model. Allow benchmarking of different model performances against industry standards or support the creation of new standards. Also, it can help support the enhancement of cybersecurity resiliency strategies. 40 2.13 Human Factors 2.13.1 Current Landscape New systems and technologies are frequently being introduced to improve railroad safety, efficiency, and productivity. New systems perform duties including CBTC, energy management, situation awareness enhancement, improving worker safety in yards and work zones via wearable technologies, remote inspections to improve defect detection, helping yard workers avoid uncontrolled movements, driver assistance, steps toward automation, etc. New technologies include AI, track or object sensors, and exoskeletons. These systems and technologies generally require human interaction for monitoring, for providing information and for handling anomalous situations. Operating procedures and human-machine interface (HMI) designs that minimize human errors in interacting with these systems can be critical to avoiding negative impacts on railroad safety and operations. Training, experience, workload, fatigue (including its relation to level of automation), and declining situational awareness of operators and other actors (e.g., highway vehicle drivers) are also relevant factors. Studies performed for railroad and other industries in other countries have produced relevant human factors information, identification of issues, probabilities, conclusions, mitigations, and analysis methodologies. In many cases, systems developed and deployed in other countries can be leveraged to save development costs and facilitate interoperability. However, differences in objectives, operating rules, environment, infrastructure, and language may prevent direct adoption. The SMEs identified opportunities for future research that fell into three general sub- themes: 1) human factors analysis methods, 2) specific topics for human factors analysis, and 3) existing solutions to leverage. 2.13.2 Opportunities for Future Research Research Opportunity Previous research in the area Outcomes of proposed research opportunity Value proposition of stated research opportunity for Canadian stakeholders The example PTC Cognitive Task Analyses (CTA) results referenced can provide a starting point for