• HOME
  • R&D
  • Major Results of Research and Development in Fiscal 2021

Major Results of Research and Development in Fiscal 2021

This publication is a collection of the major results obtained from research undertaken by the Japanese public body, the Railway Technology Research Institute, in 2021.
We would like to take this opportunity to thank all JR companies in particular and all the research centers, universities, industrial partners and other entities which have offered their cooperation, commitment and support to achieve these results and without whom none of this work would have been possible.

Ⅰ. Improvement of Safety

1. Earthquake early warning method for the earthquake source fault regions

  • We developed an earthquake warning method that reflects the amplification characteristics of seismic motions at different locations and issues a warning when the P-wave exceeds a threshold specified.
  • S-wave, which represents the principal motion, is predicted in real-time by multiplying the observed Pwave by the S-wave/P-wave amplitude ratio at various locations prepared in advance. A warning is issued when the ratio exceeds a threshold specified.
  • A warning signal can be output in as little as one second or even less in the earthquake source fault regions.

  • We developed an earthquake warning method that reflects the amplification characteristics of seismic motions at different locations and issues a warning when the P-wave exceeds a threshold specified.
  • S-wave, which represents the principal motion, is predicted in real-time by multiplying the observed Pwave by the S-wave/P-wave amplitude ratio at various locations prepared in advance. A warning is issued when the ratio exceeds a threshold specified.
  • A warning signal can be output in as little as one second or even less in the earthquake source fault regions.

2. Design method for bearing parts to control damage and improve restorability

  • We proposed a design method for bearing parts that increases the precision of calculations of the strength of parts with embedded stoppers and limits damage to parts that are difficult to restore after an earthquake.
  • The new method reduces construction costs of hard-to-restore damaged parts after an earthquake by 80%.

  • We proposed a design method for bearing parts that increases the precision of calculations of the strength of parts with embedded stoppers and limits damage to parts that are difficult to restore after an earthquake.
  • The new method reduces construction costs of hard-to-restore damaged parts after an earthquake by 80%.

3. Method for verifying the restorability of railway structures considering small-and medium-scale earthquakes

  • We proposed a method for verifying the restorability of railway structures considering small-and medium-scale earthquakes as well as large-scale earthquakes (Level 2 seismic motion)
  • With the method, the number of days needed for restoration after an earthquake is used as a verification index in the design of new structures and the assessment of existing structures, making it possible to construct structures that are easier to restore.

  • We proposed a method for verifying the restorability of railway structures considering small-and medium-scale earthquakes as well as large-scale earthquakes (Level 2 seismic motion)
  • With the method, the number of days needed for restoration after an earthquake is used as a verification index in the design of new structures and the assessment of existing structures, making it possible to construct structures that are easier to restore.

4. Method for evaluating vehicle safety in strong winds using probabilistic risk assessment

  • We proposed a method to evaluate the safety of vehicles based on the probability of external wind force exceeding the overturning resistance of vehicles to strong winds by expressing both the overturning resistance of vehicles to strong winds and external wind force as a probability distribution.
  • The method can be used to set wind speed values for operation control in accordance with critical wind speeds of overturning for vehicles and actual wind conditions.

  • We proposed a method to evaluate the safety of vehicles based on the probability of external wind force exceeding the overturning resistance of vehicles to strong winds by expressing both the overturning resistance of vehicles to strong winds and external wind force as a probability distribution.
  • The method can be used to set wind speed values for operation control in accordance with critical wind speeds of overturning for vehicles and actual wind conditions.

5. Inspection of gaps by hammering slab track filling layers and prediction of deterioration

  • We developed a method for detecting gaps between track slabs and filling layers by combining hammering tests and machine learning, as well as a test device capable of inspecting one track slab every five minutes.
  • We also developed a method for predicting strength loss in filling layers and gaps between track slabs and filling layers due to frost damage and other causes, by means of a coupling analysis of heat conduction and structure.

  • We developed a method for detecting gaps between track slabs and filling layers by combining hammering tests and machine learning, as well as a test device capable of inspecting one track slab every five minutes.
  • We also developed a method for predicting strength loss in filling layers and gaps between track slabs and filling layers due to frost damage and other causes, by means of a coupling analysis of heat conduction and structure.

6. Method for estimating snow accretion and snow dropping of a train vehicle for investigating countermeasures against snow dropping

  • The amount of snow that accumulates on a bogie and the locations where snow drops off from it can be estimated in real time from weather information along the line.
  • The growth of snow accretion on the cover plate of a bogie can be estimated within an error of approximately 3 cm, and sections of line where snow dropping occurs frequently can be estimated within an error of approximately 2 km.
  • This method can be used to determine whether snow removal work is necessary and to examine sections where measures against snow dropping are needed.

  • The amount of snow that accumulates on a bogie and the locations where snow drops off from it can be estimated in real time from weather information along the line.
  • The growth of snow accretion on the cover plate of a bogie can be estimated within an error of approximately 3 cm, and sections of line where snow dropping occurs frequently can be estimated within an error of approximately 2 km.
  • This method can be used to determine whether snow removal work is necessary and to examine sections where measures against snow dropping are needed.

7. Method of applying image processing-based object detection and obstacle determination to safety applications

  • We developed a method of self-diagnosing camera and image processing unit failures.
  • Applying this method to an abnormality detection equipment at level crossings, we confirmed that it is capable of fault diagnosing at 0.1-second intervals.
  • This method enables the implementation of image processing-based object detection and obstacle determination functions in safety-related equipment.

  • We developed a method of self-diagnosing camera and image processing unit failures.
  • Applying this method to an abnormality detection equipment at level crossings, we confirmed that it is capable of fault diagnosing at 0.1-second intervals.
  • This method enables the implementation of image processing-based object detection and obstacle determination functions in safety-related equipment.

8. Support system for preventing a decline in alertness level of drivers

  • We developed a system that detects the alertness level of drivers with images and issues an alarm.
  • It can estimate alertness level with an accuracy of 80% or higher, even when the driver is wearing a mask
  • With the proposed alarm presentation method, 90% of people with reduced alertness level reported that they felt “awakened.”

  • We developed a system that detects the alertness level of drivers with images and issues an alarm.
  • It can estimate alertness level with an accuracy of 80% or higher, even when the driver is wearing a mask
  • With the proposed alarm presentation method, 90% of people with reduced alertness level reported that they felt “awakened.”

9. Phased array ultrasonic testing method for easy detection of flaws in bogie frame welds

  • We formulated an ultrasonic testing method and procedure for bogie frames based on the phased array method that can easily detect flaws in welds.
  • The method easily distinguishes between flaw signals and noise, and effectively detects inclined flaws and flaws in coatings.
  • Detection sensitivity is equal to or better than that of conventional ultrasonic testing methods.

  • We formulated an ultrasonic testing method and procedure for bogie frames based on the phased array method that can easily detect flaws in welds.
  • The method easily distinguishes between flaw signals and noise, and effectively detects inclined flaws and flaws in coatings.
  • Detection sensitivity is equal to or better than that of conventional ultrasonic testing methods.

10. Oil for axle-box for Shinkansen with excellent low-temperature performance and maintainability

  • With the expansion of the Shinkansen network into extremely cold regions of Japan, we developed a lubricating oil for axle journal bearings of Shinkansen train that retains its fluidity even at temperatures below -40°C and that prevents reddening over time.
  • The lubricant has been confirmed to have excellent durability in a bench test on actual axle journal bearings for the equivalent of 800,000 km of travel.

  • With the expansion of the Shinkansen network into extremely cold regions of Japan, we developed a lubricating oil for axle journal bearings of Shinkansen train that retains its fluidity even at temperatures below -40°C and that prevents reddening over time.
  • The lubricant has been confirmed to have excellent durability in a bench test on actual axle journal bearings for the equivalent of 800,000 km of travel.

11. Seat surfaces for minimizing injuries to passengers in the event of a collision

  • We developed a seat surface for walkover seat that minimize injuries to passengers in the event of a collision.
  • To ensure comfort, a cushioning structure is incorporated into the seat surface. No major design changes to current seats are required, since the cushioning effect is achieved by the seat surface alone.
  • In the event of a railway crossing accident, the new seat surface can reduce lower limb injuries in passengers by 60% compared to the previous one.

  • We developed a seat surface for walkover seat that minimize injuries to passengers in the event of a collision.
  • To ensure comfort, a cushioning structure is incorporated into the seat surface. No major design changes to current seats are required, since the cushioning effect is achieved by the seat surface alone.
  • In the event of a railway crossing accident, the new seat surface can reduce lower limb injuries in passengers by 60% compared to the previous one.

Ⅱ. Cost Reduction

12. System for determining the degree of deterioration of wooden sleepers using images of railway tracks in front of trains

  • We developed a low-cost system for determining the degree of deterioration of wooden sleepers based on a deep learning technique that uses images of the view in front of a train captured with an off-the-shelf video camera.
  • The system can automatically judge the degree of wooden sleeper deterioration, according to a four-level classification scheme, with an accuracy of 90% or higher.
  • Since the system displays the rated degree of deterioration directly on images, defective sleepers can be visualized easily using a viewer.

  • We developed a low-cost system for determining the degree of deterioration of wooden sleepers based on a deep learning technique that uses images of the view in front of a train captured with an off-the-shelf video camera.
  • The system can automatically judge the degree of wooden sleeper deterioration, according to a four-level classification scheme, with an accuracy of 90% or higher.
  • Since the system displays the rated degree of deterioration directly on images, defective sleepers can be visualized easily using a viewer.

13. A method for constructing communication-based train control systems using public communication network

  • We proposed a design procedure for creating train control systems using public communication network such as 5G and a method for integrating information security functions in safety equipment.
  • The method enables a reduction of up to 50% in the total number of ground equipment that make up the system and promises to achieve labor-saving in tasks related to facility design and maintenance.

  • We proposed a design procedure for creating train control systems using public communication network such as 5G and a method for integrating information security functions in safety equipment.
  • The method enables a reduction of up to 50% in the total number of ground equipment that make up the system and promises to achieve labor-saving in tasks related to facility design and maintenance.

14. Automatic anomaly detection using vehicle condition monitoring data without additional sensors

  • We proposed a method for automatically detecting anomalies in vehicle equipment in operations without additional sensors using data from vehicle condition monitoring devices and machine learning driven by neural networks.
  • We confirmed that the method is capable of promptly detecting actual failures occurred in operational vehicles.

  • We proposed a method for automatically detecting anomalies in vehicle equipment in operations without additional sensors using data from vehicle condition monitoring devices and machine learning driven by neural networks.
  • We confirmed that the method is capable of promptly detecting actual failures occurred in operational vehicles.

15. An soundness evaluation index and method for dealing with ballasted track that is contaminated with soil

  • We proposed an index for evaluating the soundness of ballast includes with soil.
  • The proposed index can be applied both to ballast that has fouling over time and ballast contaminated with soil due to heavy rain.
  • We developed a Low-strength stabilization to limit the subsidence of ballasted track bed containing soil. Construction costs can be reduced by 90% compared to ballast replacement.

  • We proposed an index for evaluating the soundness of ballast includes with soil.
  • The proposed index can be applied both to ballast that has fouling over time and ballast contaminated with soil due to heavy rain.
  • We developed a Low-strength stabilization to limit the subsidence of ballasted track bed containing soil. Construction costs can be reduced by 90% compared to ballast replacement.

16. Method for reducing the frequency of maintenance on standard longitudinal sleepers at boundaries with structures

  • We proposed a method and standard structure for laying longitudinal sleepers at structure boundaries and verified the effectiveness of longitudinal sleepers in reducing longitudinal level irregularity in a test installation on a commercial line.
  • We confirmed that the method reduced the progression of longitudinal level irregularity to about 20% of that with conventional sleepers, making it unnecessary to perform track maintenance three to four times a year.

  • We proposed a method and standard structure for laying longitudinal sleepers at structure boundaries and verified the effectiveness of longitudinal sleepers in reducing longitudinal level irregularity in a test installation on a commercial line.
  • We confirmed that the method reduced the progression of longitudinal level irregularity to about 20% of that with conventional sleepers, making it unnecessary to perform track maintenance three to four times a year.

17. Method for evaluating the effect of impact load on fatigue of steel bridges at rail joints

  • The high-frequency vibration of members caused by the impact load of wheels passing over a rail joint contributes significantly to the initiation of fatigue cracks in steel bridges.
  • We developed an analytical method for evaluating cumulative fatigue that takes into account high-frequency vibration.
  • The method makes it possible to identify areas of high accumulated fatigue, thereby increasing the efficiency of inspection and reinforcement.

  • The high-frequency vibration of members caused by the impact load of wheels passing over a rail joint contributes significantly to the initiation of fatigue cracks in steel bridges.
  • We developed an analytical method for evaluating cumulative fatigue that takes into account high-frequency vibration.
  • The method makes it possible to identify areas of high accumulated fatigue, thereby increasing the efficiency of inspection and reinforcement.

Ⅲ. Harmonization with the Environment

18. Efficient method for adjusting openings in large tunnel entrance hoods

  • We proposed an efficient method for uniformly adjusting the area of windows arranged consecutively in a tunnel entrance hood.
  • Compared to the conventional method of adjusting the position of openings, the new method makes it possible to reduce micro-pressure waves by the same degree, using 80% fewer trials for adjustment.

  • We proposed an efficient method for uniformly adjusting the area of windows arranged consecutively in a tunnel entrance hood.
  • Compared to the conventional method of adjusting the position of openings, the new method makes it possible to reduce micro-pressure waves by the same degree, using 80% fewer trials for adjustment.

19. High-resolution sound source identification method for wayside noise reduction measures

  • The adoption of a new specially designed two-dimensional spiral array system and the application of a cutting-edge signal processing method for moving sound sources make it possible to increase the spatial resolution of sound source distribution around vehicles by a factor of about nine.
  • Since the new system is capable of visualizing sound source distribution with a spatial resolution of about the size of a wheel, it can be used to analyze sound sources in bogies, pantographs, and other parts in detail. It is also useful for verifying the effectiveness of noise reduction measures.

  • The adoption of a new specially designed two-dimensional spiral array system and the application of a cutting-edge signal processing method for moving sound sources make it possible to increase the spatial resolution of sound source distribution around vehicles by a factor of about nine.
  • Since the new system is capable of visualizing sound source distribution with a spatial resolution of about the size of a wheel, it can be used to analyze sound sources in bogies, pantographs, and other parts in detail. It is also useful for verifying the effectiveness of noise reduction measures.

Ⅳ. Improvement of Convenience

20. Method of evaluating the effectiveness of investing in disaster countermeasures for freight transport

  • We developed a method for quantitatively evaluating the effectiveness of investments in disaster countermeasures, using the volume of freight traffic and the costs to operators during periods when services are partially interrupted due to a disaster as indices for evaluation.
  • The method can be used for formulating disaster countermeasures and identifying investment priorities, i.e., the areas that most need countermeasures.

  • We developed a method for quantitatively evaluating the effectiveness of investments in disaster countermeasures, using the volume of freight traffic and the costs to operators during periods when services are partially interrupted due to a disaster as indices for evaluation.
  • The method can be used for formulating disaster countermeasures and identifying investment priorities, i.e., the areas that most need countermeasures.

21. Method of evaluating thermal comfort of solar radiation environments in commuter trains

  • We proposed a method for evaluating the comfort of thermal environments in commuter trains, taking into account the effects of solar radiation.
  • We conducted an experiment with participants by simulating a solar radiation environment inside a train, to confirm that the predictions generated by the proposed method correspond closely to actual experience.

  • We proposed a method for evaluating the comfort of thermal environments in commuter trains, taking into account the effects of solar radiation.
  • We conducted an experiment with participants by simulating a solar radiation environment inside a train, to confirm that the predictions generated by the proposed method correspond closely to actual experience.

Ⅴ. Basic Research

22. Earthquake early detection method based on deep learning of seismic motion

  • We developed an earthquake early detection method based on deep learning of seismic motions.
  • The method improves the ability to discriminate between seismic motion and noise from approximately 90% to 99% and enhances the estimation precision of all seismic parameters, including epicentral distance and magnitude.
  • We confirmed that the method is capable of processing estimates in real time.

  • We developed an earthquake early detection method based on deep learning of seismic motions.
  • The method improves the ability to discriminate between seismic motion and noise from approximately 90% to 99% and enhances the estimation precision of all seismic parameters, including epicentral distance and magnitude.
  • We confirmed that the method is capable of processing estimates in real time.

23. HILS for the current collection systems using a High-Speed Test Facility for Pantograph/OCL Systems

  • Using a High-Speed Test Facility for Pantograph/OCL Systems, we developed a HILS for the current collection systems that takes into account the effects of overhead contact line (OCL) vibration, sliding due to travel, stagger of the OCL, environmental temperature, humidity, and energization.
  • The HILS for the current collection systems enables comprehensive evaluation of pantograph performance based on a bench test, making it possible to develop higher-performance pantographs more efficiently and inexpensively

  • Using a High-Speed Test Facility for Pantograph/OCL Systems, we developed a HILS for the current collection systems that takes into account the effects of overhead contact line (OCL) vibration, sliding due to travel, stagger of the OCL, environmental temperature, humidity, and energization.
  • The HILS for the current collection systems enables comprehensive evaluation of pantograph performance based on a bench test, making it possible to develop higher-performance pantographs more efficiently and inexpensively

24. Mechanism of increasing wear rate of contact wires near pantograph stopping positions of Shinkansen

  • An investigation within a Shinkansen station yard clarified the train-speed dependency of contact wire wear rate.
  • The study revealed that the coefficient of friction increases at low speeds, thereby accelerating “delamination wear,” which is the cause of the high wear rate.
  • We confirmed that the application of grease can reduce contact wire wear in low-speed sections of railway lines.

  • An investigation within a Shinkansen station yard clarified the train-speed dependency of contact wire wear rate.
  • The study revealed that the coefficient of friction increases at low speeds, thereby accelerating “delamination wear,” which is the cause of the high wear rate.
  • We confirmed that the application of grease can reduce contact wire wear in low-speed sections of railway lines.

25. Text analysis codes for human error analysis

  • We created an algorithm (code) for text analysis to analyze work characteristics, psychological factors, and error behaviors described in risk information from accidents and potential incidents.
  • The code can analyze 95% of approximately 4,000 items of the descriptive data of error behaviors in a few seconds, whereas it would take a human several months to classify them.

  • We created an algorithm (code) for text analysis to analyze work characteristics, psychological factors, and error behaviors described in risk information from accidents and potential incidents.
  • The code can analyze 95% of approximately 4,000 items of the descriptive data of error behaviors in a few seconds, whereas it would take a human several months to classify them.

26. Method of visualizing sound wave propagation for clarification and prediction of railway noise

  • We developed a measurement method for visualizing sound waves propagating through the air based on laser polarization and used it to visualize sound waves diffracting through sound barriers and the Doppler effect caused by a moving sound source.
  • We developed a method for estimating the sound pressure distribution in a specific cross-section based on sound wave visualization results.
  • The method can be used to shed light on the phenomenon of railway noise and increase the accuracy of noise prediction methods along railway lines.

  • We developed a measurement method for visualizing sound waves propagating through the air based on laser polarization and used it to visualize sound waves diffracting through sound barriers and the Doppler effect caused by a moving sound source.
  • We developed a method for estimating the sound pressure distribution in a specific cross-section based on sound wave visualization results.
  • The method can be used to shed light on the phenomenon of railway noise and increase the accuracy of noise prediction methods along railway lines.

·I.2, I.3 received funding from the Ministry of Land, Infrastructure, Transport and Tourism's subsidy aid of railway technology development.
·I.5 is collaborative research with Tohoku University and University of Yamanashi.
·I.7 is collaborative research with Kyosan Electric Mfg. Co., Ltd..
·I.10 is collaborative research with ENEOS Holdings, Inc..
·II.15 is collaborative research with Waseda University.
·V.26 is collaborative research with Kyoto University.

Major Results of Research and Development