Automated track inspection is designed to help railroads tackle potential pitfalls, but what happens when track inspectors run into challenges?
On a perfect day, service providers of automated track inspection will have generous work windows, unobstructed views of the track and zero false positives. But, if they were all perfect days, this wouldn't be railroading. It's up to these service providers to get the job done whether obstacles present themselves or not. Today's inspection technologies offer custom parameters, can operate at higher speeds and provide better data for a complete view of track health.
Balfour Beatty Rail
Balfour Beatty Rail, Inc., holds the contract for maintenance along the Alameda Corridor, which includes track, signal, communications and right-of-way maintenance. As Larry Mahon, area operations manager for Balfour Beatty Rail, says, one of the biggest challenges facing track inspection along the corridor is allotting enough time to perform track inspections.
"On some days, the heavy volume of traffic makes thorough and complete inspections more difficult than usual," said Mahon. "The Alameda Corridor is a three-mainline corridor. Because of Federal Railroad Administration (FRA) regulations, we can't inspect all three tracks at once unless there's a second qualified inspector in the vehicle and the track centers are no more than 39 feet apart. On the Alameda Corridor, the track centers are 15 feet apart, but there are 143 power switches and each has to be inspected monthly. With traffic the way it is, it's a challenge making thorough track inspections and walking to each switch.
"There's no substitute for good visual track inspections. Geometry cars will find things the eye may miss, but they can't find things that a walking person will. It all comes down to working with the dispatching center to make sure everyone understands the requirements so the FRA regulations are met," said Mahon."
Two main challenges surround track inspection, according to ENSCO, Inc. The first is scheduling critical track inspection activities with limited track time without impacting revenue service. The second is identifying rail breaks prior to a derailment.
"With high traffic volumes on freight lines and the growth of passenger services around the country, as well as ever increasing demands placed on railroad workforces, the time available for the range of various inspections required is becoming more difficult to schedule," said Eric Sherrock, senior staff engineer at ENSCO. "With regards to track components, the identification of rail breaks prior to a derailment is of critical importance to the industry, especially with the impending implementation of positive train control and the eventual decrease in reliance on track circuits. Rail breaks continue to be identified as a leading cause of a significant number of track-related derailments."
ENSCO notes that recent procurements by railroads indicate a trend towards comprehensive track inspection in a single pass from vehicles equipped with several complimentary inspection systems. The company's range of track inspection systems can be integrated on a single vehicle to provide comprehensive automated track inspections. In 2012, ENSCO delivered a full-sized self-propelled inspection vehicle to Washington Metropolitan Area Transit Authority capable of simultaneously measuring track geometry, rail profile, platform edges, as well as third rail geometry and temperature. The vehicle is also able to record video and thermal images of the tunnels and conduct ultrasonic rail-flaw inspections, all while capturing foot-by-foot track measurements in a single pass.
ENSCO's Comprehensive Track Inspection Vehicle (CTIV) is a hi-rail based platform that allows for track geometry and rail profile measurement along with machine vision-based assessment of joint bars and track bed components including ties and fasteners.
ENSCO says it also offers a suite of autonomous inspection systems that provide high quality inspection results from revenue service vehicles with minimal direct involvement from operators.
"One of the most significant advantages of autonomous inspection technology is that every movement of the host train offers an opportunity to evaluate the track, allowing for more frequent inspections without track time being consumed by dedicated inspection vehicles. The use of autonomous inspection technologies can result in earlier detection of track defects, allowing for maintenance practices to be preventative rather than reactive, ultimately reducing the number of track related derailments," said Sherrock.
ENSCO provides both the Vehicle/Track Interaction (V/TI) Monitor and the Autonomous Track Geometry Measurement System (ATGMS). Currently, ENSCO has more than 250 autonomous V/TI Monitors in operation and delivered two ATGMS units this year, including the first ever to be deployed by a Class 1 railroad.
"Autonomous V/TI Monitors have the ability to capture locations of a broken rail by measuring the response of a vehicle's unsprung mass to the track and reporting the locations of severe impacts to end users. ENSCO's machine vision-based inspection systems are able to capture high-resolution images of track components, analyze these images for cracks and breaks and alert users to their existence. Both of these approaches can be an effective means of reducing the risk of derailments due to rail breaks," said Sherrock.
"In 2005, railroads dismissed as novelty, the thought of using machine vision technology for tie inspection and grading," said Lynn Turner, vice president marketing and sales for Georgetown Rail Equipment Company (GREX). "Today, GREX's Aurora Automated Tie Inspection system is widely accepted among America's railroads as a planning tool for tie replacement and deterioration trending models. Using the technology of high-speed cameras and lasers, Aurora is state-of-the-art tie condition inspection and reporting."
Turner continued, "GREX develops exclusive tie grading models for each customer using sophisticated algorithms that embed unique grading characteristics specific to that customer. These unique characteristics are identified during a walking calibration process with the human tie inspector. Aurora then 'learns' these grading attributes and begins to mimic the expert inspector, mirroring the grading results. This output, in a 3-D digital track image along with the statistical analysis of tie condition, allows critical and thorough reporting and assigns each tie an exact GPS location signature. Incorporating rolling 39-foot defects, joint tie defects and clusters of failed ties into customized reports are typical of customer presented challenges that GREX engineers have developed solutions for. It is no coincidence that customers who have invested the most effort into Aurora have realized the best performance."
Just as a walking inspector requires a clear view of the track to perform at an optimum level, so does Aurora in order to make an accurate assessment of tie condition. Turner says keeping that view free of obstructions, such as ballast covered ties, is a common challenge.
"Originally, Aurora was unable to distinguish these ballasted ties. With ballast coverage being one of the most prevalent reasons for track obstruction, GREX focused on improving Aurora's ability to better handle these instances. When Aurora now encounters ballast covered ties, it has been taught to recognize the obstructing ballast and identify the areas of ties that are still visible. Using a percentage of visibility calculation, Aurora can provide an assessment based on what it can see. Plate cut can also be measured and made a part of the overall assessment," said Turner.
"Every challenge that Aurora overcomes has a positive customer impact. Each time, the system's robustness is improved and the quality of data is refined. Our customers give testimony to the progress Aurora has made and are confident in our inspection abilities, deploying Aurora in areas that were once deemed too obstructed or unknown to inspect with automation. Aurora engineers will continue to solve the challenges that customers identify but Aurora's acceptance has proven that it is here to stay, providing accuracy and repeatability at more than 40 miles per hour, day or night," said Turner.
Joe Palese, senior director, engineering & technology at Harsco Rail, notes that track inspection needs to take into account safety and maintenance.
"Thus, inspection is first focused around those critical components and areas that lend themselves to decreasing the risk of failure and, subsequently, increasing safety. The data gathered during the inspection of any component can also be turned into a wealth of information for understanding component degradation and failure. This allows for developing maintenance practices to prolong the life of the components. Addressing the most critical components first is of paramount importance. This needs to be accomplished with an effortto keep costs as low as possible," said Palese.
Palese points to Harsco Rail's Automated Switch Inspection Vehicle (ASIV) as one approach to incorporating safety and maintenance into track inspection. The company says the ASIV is a hi-rail based turnout inspection system that utilizes proven inspection technology to focus on the rail and geometry portions of the turnout, which allows for cost effective inspection of the critical safety components of the turnout.
"Traditional methods are used to inspect the remaining portions. Once cost effective production measurements systems become available for these remaining portions, they will be added to the ASIV. Research is currently under way to address this," said Palese.
Another issue Palese mentions regarding inspection is speed.
"With track time being limited, most customers require inspection to occur at track speeds. This is sometimes not possible with current inspection technologies. Customers want inspection at track speed and for it to be autonomous as opposed to dedicated inspection," he said. "Harsco Rail has been investigating and developing technologies that can be deployed on hi-rail vehicles for easy on/off to take advantage of short work windows. In addition, Harsco Rail has been investigating deploying these systems on other work equipment (grinders, tampers and UTVs) to take advantage of the allocated work window."
Holland LP says a challenge in track measurement is the validation of rail cant measured by track geometry test vehicles.
Holland's TrackSTAR® fleet measures rail cant along with track geometry in one foot increments. Bob Madderom, vice president and general manager with Holland LP, says it hasbecome standard practice for railroads to measure and report on rail cant exceptions from automated track geometry cars.
"Essentially, we now have the capability to identify rail cant exceptions in the normal course of our reporting for all measured track," said Madderom. "Rail cant measurement has become a very important measurement of condition of track. Differential plate cutting on wood ties in heavy tonnage or poorly maintained track can be a problem. Also, rail cant variation on concrete ties is an effective way to identify potential areas of worn pads, or worse, rail seat deterioration (RSD).
Madderom says rail cant variation is effective in identifying RSD on concrete ties and points to recent FRA track safety standards (TSS 213.234) that became effective in July 2012 and require the automated measurement of concrete tie track for indications of potential RSD.
"Although geometry cars have effective systems to measure rail cant and provide exception reporting, until recently, there have been no manual inspection tools to easily measure existing condition cant in track. Field inspectors were dependent on mostly visual inspection and GPS or MP + footage information from reports to try to find cant problems," said Madderom. "Recently, Holland completed the design of a new light-weight rail cant measurement tool (RCMT). Similar to the inspector's track gauge measurement tools in weight and size, the patented RCMT can provide direct read out of existing rail cant in a few seconds. The RCMT eliminates the guesswork of where rail cant exceptions begin and end and provides immediate feedback to the track maintenance person."
"The biggest challenge for our customers in track inspection is having a means to measure vertical track deflection under heavy-axle-loads, which gives them an indication of track formation condition," said G. Robert Newman, business development engineer for TÜV Rheinland Rail Sciences, Inc., which partnered with MRail on a system to take on this challenge.
"Railroads need to quantify this value over sections of track at a normal train speed and they need to make this measurement often in order to determine the trends. This in turn gives them a fast, simple method to evaluate the condition of the track formation to plan further inspection and track maintenance planning," said Newman.
"The new MRail System, provides a simple measurement in inches, indicating total vertical track deflection under heavy-axle-load. The measurement can be made unattended and at normal train speed for each track section. Because the data is recorded and transmitted remotely at a high speed on a revenue train, no track time is required. This allows the railroad to take frequent measurements to establish trends in formation condition," said Newman.
NxGen Rail Services
Robert Grant, director of engineering sales for NxGen Rail Services, a Sasser Family Holdings Company, believes that as railroad maintenance activities continue their evolution from reactive to proactive, condition based maintenance of track and rolling stock will provide significant savings, as well as improvements in safety for the railroads in the future.
"The challenge for the track inspection industry is therefore to provide the quality data required to meet this step change," said Grant. "Accurate technologies are now available that can be used, or combined, to test and measure almost all of the required parameters and components. The issue is to understand how to combine and apply them to provide the information in the format that the industry needs.
"At NxGen Rail Services, we have met this challenge by integrating the different technologies onto a common platform that travels at line speed, providing the maintenance planners with a holistic view of their asset in near real time. This provides a fast, non-intrusive solution. Speed is also necessary to be able to cover the huge mileage of track that make up the rail networks with an adequate frequency, but with a minimum level of investment in systems. However, more speed generates a greater frequency of defect detections, so our systems have also been tailored to use machine intelligence to understand which defects are important and generate accurate reports," said Grant.
Grant says the system gives precise locations of defects, enabling planners to examine them from different points of view to better understand the causes of the problems and design better solutions.
"For example, a cluster of missing fasteners might be the result of a geometry defect or dipped joint, caused by an inadequate layer of ballast sitting on an unstable track bed or a bad transition from one rail profile to another or a combination of both. Multiplexing technologies to understand the root cause means that the repair can be designed to resolve the problem definitively, whereas just replacing the fasteners and tamping the track would only have a temporary effect, that will need to be repeated again and again as the consequence of the underlying defect continues to reappear," said Grant.
Tom Keogh, president of Rail Radar, recognizes that while track inspection is essential, it consumes both financial and human resources and sometimes more importantly, track capacity.
"One of the largest issues dealing with inspection of track is that of acquiring sufficient time on the track to conduct detailed surveys. One of the main issues is most areas in most need of inspection are ones experiencing highest traffic densities. Conventional tie assessment requires either walking tie inspectors and/or image based systems to assess ties and related components. These methods are challenged by subjective interpretations and/or features, which obscure analysis (color changes, debris, ballast, etc.) and can be incorrectly identified as defects."
Rail Radar has developed a three-dimensional (3-D) tie assessment system that the company says improves upon existing 2-D technologies. According to Rail Radar, the 3-D tie assessment results show improved in-situ tie condition data (dimensional correct 3-D tie cross-sections); allows data to be collected at increased speed; has enhanced tie processing with improved quality assessment methods and algorithms; has improved unambiguous feature identification, which reduce false positives; incorporates vertical height information to improve feature identification and classification and has enhanced feature classification algorithms that use 3-D feature models to improve reliability and speed.
"Technology has evolved and the benefits of incorporating automation through machine vision into inspection processes are well documented and proven. Rail Radar's 3-D system uses industry proven 'machine vision' techniques to improve the efficiency, effectiveness and objectivity of the track inspection process. Rail Radar's 3-D system capitalizes proven technology first through data acquisition, second through image analysis and third through data analysis. Rail Radar has a proven 'track' record working with major railroads to ensure experienced tie inspection processes and thresholds are upheld. The enhanced quality of data collected with Rail Radar's 3-D system ensures this happens," said Keogh.