TRACK geometry is considered the foundational measurement for assessing track condition. Through the use of dedicated manned survey vehicles, automated track geometry measurement has become a cornerstone of the railway industry’s safety programme for more than 40 years. Autonomous track geometry measurement systems (ATGMS) represent the latest technological advance in geometry measurement and are already producing significant results.
ATGMS collects the standard track geometry measurement parameters without direct operator involvement, and since provisions for onboard operators are not needed, an ATGMS can be installed on a revenue service vehicle. By reducing the reliance on specialised inspection vehicles, ATGMS technology reduces the life cycle cost of geometry measurement operations and virtually eliminates interference with revenue service operations. These features allow an ATGMS to survey significantly more track at a lower cost than traditional manned track geometry inspection vehicles.
In the United States, ATGMS was first deployed by Ensco on behalf of the Federal Railroad Administration’s Office of Research, Development and Technology (FRA RD&T). The two organisations deployed autonomous systems of varying capabilities as part of a technology development plan intended to guide research while demonstrating the benefits of autonomous track evaluation.
The first deployment of ATGMS commenced in January 2008 when the FRA and Ensco implemented a prototype, bogie-mounted track geometry measurement system on a passenger coach operated within Amtrak’s Auto Train service using CSX’s network.
The prototype ATGMS recorded track geometry data and sent reports of events exceeding FRA class-based track geometry thresholds to stakeholders for evaluation and eventual follow-up on track. Through a cooperative agreement between the FRA, CSX and Amtrak, the first ATGMS was allowed to operate from January 2008 to March 2011, surveying almost 740,000km of track. This extensive testing led to improvements in robustness and reliability.
The subsequent use of ATGMS centred on demonstration in widespread revenue service as part of the FRA Office of Safety’s Automated Track Inspection Program (ATIP). ATIP employs a fleet of FRA-owned track inspection vehicles to assess track safety in the US. Several technical capabilities were deployed, most notably the transmission of all sensor data collected by the system via public cellular networks to a remote server for processing. This was the first time continuous track geometry data was recorded and transmitted to stakeholders in near real time.
Concurrent with this testing, a secure internet-accessible remote editor desk console was employed by a trained operator for near real-time review and validation of survey results including exceptions to class-based track geometry limits.
In 2012, the FRA and Ensco deployed a carbody-mounted autonomous track geometry system on an Amtrak 202km/h Amfleet coach. The movement of the system to the carbody increased system reliability by reducing vibration and increasing protection from flying debris. The new configuration also minimised interference during bogie and wheelset maintenance. The ability to install the system on a carbody facilitates deployment of ATGMS on a wide range of vehicle designs and reduces design, construction, installation, and maintenance costs compared with bogie-mounted equipment.
The Amfleet coach operated in revenue service on Amtrak’s Northeast Corridor between Washington, DC and Boston. The ATGMS transmitted data to the central processor in near real-time, using the local cellular network and produced accurate continuous track geometry measurements. Analysis of data collected during this demonstration illustrated that the carbody-mounted ATGMS produced data of equal quality to that of a bogie-mounted track geometry measurement system.
The FRA continues to increase its utilisation of ATGMS technology. FRA modified its DOTX 216 inspection car to collect data in both manned and autonomous modes. In 2015, the FRA’s ATIP ATGMS programme surveyed more than 68,000km of track, or 55% of the total surveyed by the ATIP fleet.
The FRA has purchased two box wagons to equip with ATGMS technology. One wagon will be used for an ATGMS demonstration on US short line and regional railroads; the demonstration wagon will be self-powered, using batteries charged by solar panels and a small fuel cell.
International projects
The concept is also in use outside of the United States. In 2012, Ensco delivered an unmanned track geometry and rail profile system to Fortescue Metals Group (FMG), an iron-ore mining operation in the Pilbara region of northwest Australia. FMG operates 620km of track between the mines and the seaport. The system was installed on a revenue Dash 9 freight locomotive. Manual download of the data was chosen because the locomotive routinely returns to the port. Similar to early FRA ATGMS operation, the data is reviewed in the office and is used in FMG’s maintenance planning for rail grinding, rail replacement, and surfacing.
In addition, Ensco delivered the first commercial ATGMS to Canadian Pacific Railway (CP) in 2013. The body-mounted ATGMS is mounted on a box wagon. The box wagon has a poured concrete floor to achieve a 26.3 tonne gross axleload. Electrical power was supplied by a hybrid solar cell, diesel generator, and battery system.
The system streams data to a remote server, but unlike other ATGMS systems, the CP ATGMS remote server was updated to remove the need for the editor desk human operator.
When data arrives at the CP ATGMS remote server, it is first evaluated to determine which track was surveyed. This is performed by using the differentially-corrected GPS data and comparing it with a track centreline basemap of the CP rail network. Additional algorithms were deployed to handle instances of GPS loss due to tunnels or tree cover.
Once the track is determined, its corresponding thresholds are applied and exceptions are identified. Automated algorithms evaluate the track geometry data associated with the exceptions to determine if the exceptions are valid or invalid.
The validated reports are then emailed to CP personnel. When the CP wagon passes a track perturbation, the entire process of transmitting the data, automatically identifying the track, automatically confirming validity, and emailing the report takes approximately five minutes.
At the 2015 American Railway Engineering and Maintenance Association (Arema) annual conference, CP presented results of their evaluation of their ATGMS in the paper “Implementation and Use of Class I Railroad Autonomous Track Geometry Measurement System.” CP tested the reproducibility of the system by operating it over the same track at different speeds and directions. Results showed that the ATGMS achieved the desired reproducibility results.
CP compared the results of the autonomous system with one of their manned vehicles by coupling them into the same train while conducting surveys. When the track geometry data from the two systems were overlaid, CP observed very good correlation. CP compared the automated track determination results with the identified tracks from the manned vehicle and found a 97% correlation. More than 4000 exceptions were reviewed manually and evaluated with the automated algorithm. Results of the manned survey and the algorithm were compared and were found to have a 96% match.
CP is currently operating the ATGMS wagon in revenue freight trains by waybilling it from point-to-point. The wagon can be placed anywhere within the train and does not require special marshalling. Generally the wagon is waybilled repeatedly over long key routes to obtain repeated measurements and ensure that all parallel main line and sidings are surveyed.
ATGMS has been shown to be a flexible, efficient tool for use in quality control and maintenance planning activities. It has been shown to reduce the life cycle costs of geometry measurement operations; eliminate interference with revenue operations; increase inspection, frequencies and productivity; and provide data of the highest quality possible. The role of autonomous inspection systems in track condition monitoring will continue to expand in the coming years due to their ability to improve safety with increased surveying capability and decreased cost.
