Project number 314408 FP7-SST-2012-RTD-1
MAXBE INTEROPERABLE MONITORING, DIAGNOSIS AND MAINTENANCE STRATEGIES FOR AXLE BEARINGS
D9.1 – Recommendations for interoperable monitoring and maintenance strategies for axle bearings
October 2015
Partners: Instituto Superior Técnico (IST) University of Porto (UPORTO) Dynamics, Structures and Rede Ferroviária Nacional (REFER) Systems International (D2S) Ansaldo STS (ASTS) Vlaamse Vervoermaatschappij Institute of Transport, Railway Construction De Lijn (DL) and Operation (IVE) Empresa de Manutenção de COMSA Equipamento Ferroviário (EMEF) Evoleo Technologies (EVOLEO) I-MOSS NEM Solutions (NEM) Krestos Limited (KRESTOS) University College Cork (UCC) University of Birmingham (UoB) MERMEC NOMADTech (NOMAD) SKF Industry
Project number 314408 FP7-SST-2012-RTD-1
Revision history Document - D9.1 – Recommendations for interoperable monitoring and maintenance strategies for axle bearings Version Date
Author
Description/Remarks/Reasons for change
0.1
2015-09-14
UPORTO
First issue
0.2
2015-09-14
UOB/KRESTOS First issue additions made
0.3
2015-10-12
IMOSS, DL and D2S
Input from De Lijn – D2S – I-moss Proposal for document restructuring
0.4
2015-10-12
IVE
Conclusions and Remarks
0.5
2015-10-14
UCC
Condition-based Maintenance Schedule model
0.6
2015-10-14
COMSA
Tool for optimal physical distribution of the monitoring systems
0.7
2015-10-15
SKF
Wireless On-board monitoring system
0.8
2015-10-19
ASTS / UNIGE
WP5 – Integration of Wayside and On-board Monitoring Systems
0.9
2015-10-29
IST
Multibody Dynamic Modelling
1.0
2015-10-30
NEM
Maintenance schedule optimisation tool; Smart Diagnostics and Information Integration
1.1
2015-10-31
UPORTO
Final review
Copyright © MAXBE Consortium
2 of 33
CONTENTS 1. INTRODUCTION ...................................................................................... 4 2. WP2 – TECHNOLOGY ASSESSMENT AND SPECIFICATION ........................... 5 2.1. LABORATORY TESTS BY UPORTO ....................................................................... 5 2.2. LABORATORY TESTS BY UOB................................................................................ 7 2.3. MULTIBODY DYNAMIC MODELLING .......................................................................... 9
3. WP3 – ON-BOARD CONDITION MONITORING SYSTEMS ............................ 11 3.1. W IRELESS ON-BOARD MONITORING SYSTEM – SKF .............................................. 11 3.2. ON-BOARD MONITORING SYSTEM DEVELOPED BY UPORTO AND EVOLEO............ 14 3.3. ON-BOARD MONITORING SYSTEM INSTALLED BY UOB ............................................ 16
4. WP4 – WAYSIDE CONDITION MONITORING SYSTEMS .............................. 19 4.1. VIBRATION MONITORING SYSTEM FOR FREIGHT, URBAN AND LONG-DISTANCE TRAINS 19 4.2. LRV WAYSIDE MONITORING ................................................................................ 21 4.3. ACOUSTIC EMISSION MONITORING SYSTEM ........................................................... 22
5. WP5 – INTEGRATION OF WAYSIDE AND ONBOARD MONITORING SYSTEMS. 26 5.1. W AYSIDE AND ON-BOARD INTEGRATION SOFTWARE .............................................. 26 5.2. TOOL FOR OPTIMAL PHYSICAL DISTRIBUTION OF THE MONITORING SYSTEMS ............ 28
6. WP7 – ASSET MANAGEMENT TECHNOLOGIES ........................................ 30 6.1. CONDITION-BASED MAINTENANCE SCHEDULE MODEL ........................................... 30 6.2. SMART DIAGNOSTIC ........................................................................................... 31 6.2.1. Smart diagnostics ................................................................................................................... 31 6.2.2. Information Integration ........................................................................................................... 31
7. RECOMMENDATIONS ............................................................................ 32
Copyright © MAXBE Consortium
3 of 33
1. INTRODUCTION The strategic objective of the MAXBE project is to develop and to demonstrate novel wayside and onboard axle bearing monitoring technology capable of detecting axle bearing faults and defects at a much earlier stage than currently possible. Also it is expected within the framework of the MAXBE project to define recommendations for the interoperable axle bearing monitoring arising from the development of the new wayside and onboard monitoring technology. The MAXBE project was accomplished by the execution of nine workpackages: - WP 1 Consortium coordination and management - WP 2 Technology assessment and specification -
WP 3 Onboard systems
- WP 4 Wayside systems - WP 5 Integration of systems - WP 6 Testing and validation of systems - WP 7 Development of asset management technologies - WP 8 Demonstration - WP 9 Dissemination This document presents a summary of the results achieved in each workpackage of the MAXBE project mentioned its impacts for the railway community. This report also indicates recommendations for interoperable monitoring and maintenance strategies for axle bearings that derived from the project results and achievements.
Copyright © MAXBE Consortium
4 of 33
2. WP2 – TECHNOLOGY ASSESSMENT AND SPECIFICATION 2.1. Laboratory Tests by UPORTO Within WP2, an exhaustive analysis of the failure modes and the associated degradation process of several types of axle bearings was performed by extending an existing cause-and-effect matrix. For defining the matrix: i) an analysis of events causing axle failure based upon data from the operators of the MAXBE project; ii) a classification of the bearings failures modes (friction, wheels flats, poor alignment); iii) an assessment of the indicators for the different failure modes (temperature, noise, etc.) were considered. Another result from the work in WP2 was the development of an interactive and friendly-user data-base in Microsoft Excel (optimized for version 2010) that allows the user to observe all the characteristic, problems, properties, etc in detail, without an information overflow.
Figure 2.1 – Interactive Data-base of axle bearing failure modes and degradation process
Furthermore, aiming the assessment of the real bearing failures and of lubricant degradation during and after service, the analysis of bearing/grease reliability and maintenance data that can be used in RAMS models and also the analysis of the correlation between real bearing operating parameters and model prediction, related to temperature, vibration, dynamic loading and grease ageing, several analysis were performed in axle bearings from PCC Antwerpen unit (tram that circulates in Antwerp) and UME 3400 unit (Urban train of Oporto). That research included: a Ferrography analysis, a viscosity analysis and a surface analysis (visual inspection, microscopic surface analysis and roughness measurements).
Copyright © MAXBE Consortium
5 of 33
(a)
(b)
(c)
(d)
Figure 2.2 - (a), (b) Roughness measurement of the damage observed in a roller (crater); (c) Microscopic Surface Analysis; (d) Visual Inspections
In order to assess the mechanisms behind the friction torque and power loss of the rolling bearings under oil or grease lubrication, rolling bearing tests were performed with a modified Four-Ball Machine. The four-ball arrangement was substituted by a rolling bearing assembly developed to test several rolling bearings and measure the friction torque and the operating temperature in several different points. The friction torque was measured with a piezoelectric torque cell and the temperature measurements were performed through thermocouples installed at strategic locations. Both types of axle bearings (PCC Antwerp and UME3400) were analyzed considering the specific load conditions and lubricants. As expected, it was verified that the operating temperature increases as the speed and load is also increasing, being the speed the most relevant parameter responsible for the temperature increase. Vibration analysis in a faultless axle bearing from the UME3400 unit was also performed considering two different load conditions: unloaded and loaded using accelerometers and a dynamic signal analyzer. The experimental set-up consists of a test rig which supports a dc servomotor with a speed controller, a shaft drive, a test bearing housing which accommodates an axel bearing and a hydraulic load applicator. The testing was carried out for the shaft running at approximately 264 rpm (4.4 Hz). The vibration signals were processed by a FFT conventional approach in the frequency domain were all vibration components including: bearing vibration, unbalance and misalignment were analyzed.
Copyright © MAXBE Consortium
6 of 33
2.2. Laboratory Tests by UoB Laboratory tests were carried out on healthy and defective bearings using a customized test rig with capable of rotating sample bearing from 100 up to 1000 Revolutions Per Minute (RPM). An R50A resonant AE piezoelectric sensor procured from Physical Acoustics Corporation (PAC) was mounted on top of the bearing case using a magnetic hold-down. The AE sensor was coupled on the bearing case using Vaseline. The AE signals were amplified using a preamplifier and amplifier also from PAC by 43dB. The AE signals were digitised using an Agilent 2531A data acquisition card (DAQ). Customised software written in MATLAB by UoB and Krestos was used to log and analyze the captured data. A universal general purpose accelerometer was also used to replace the acoustic emission sensor in each test. Figure 2.3 shows the test rig used in this experiment.
Accelerometer
Bearing
AE sensor
Figure 2.3 - Laboratory bearing test rig in Gisbert Kap building, University of Birmingham.
The bearing samples used in the laboratory rig tests were PFI Inc, model PW29530037CSHD Ford wheel bearing with dimensions of 28 x 53 x 37 mm. These bearings were disassembled; defects were induced and put back on the rig and include a healthy, outer race and roller defective. Figure 2.4 shows an example of the bearings.
Figure 2.4 - Example of a rig double-raw tapered roller bearing
Copyright © MAXBE Consortium
7 of 33
The overall results showed that: the crest Factor does not seem to be a suitable method to analyse the vibration data; the Kurtosis is recommended for the vibration applications. The main drawback of the Kurtosis is that it begins to revert back to the undamaged value as the speed increases. Crest Factor shows very poor results in both vibration and AE data. Laboratory experiments were also carried out to simulate real conditions using a test trolley. Figure 2.5 shows a motorized trolley carrying a test wheel (wheel diameter 0.16m) with a metal build-up defect on the 7m long test track at the University of Birmingham.
Metal Built-up defect
AE sensor
Figure 2.5 - Experimental work with motorised trolley.
Copyright © MAXBE Consortium
8 of 33
2.3. Multibody dynamic modelling The monitoring of the axle bearing health requires the knowledge of the frequency response of the healthy bearings as a necessary data for the understanding of any abnormal condition. It happens that not only such data is unavailable, due to the fact that the bearing manufacturers do not provide it, but also the information required to address the defect signatures of the axle bearings and their association with specific defects is not available, for the same reasons. The alternative to obtain this fundamental response of the axle bearings is by developing an extensive program of experimental measurement of the axle bearings, to obtain their geometric, material and tribological characteristics, to identify the geometric, material and tribological characteristics of the defects. The second step to develop the necessary tools to obtain reliable and fundamental reference dynamic response of the axle bearings in realistic operation conditions requires that computational tools to address the dynamic analysis of axle bearings in realistic operational conditions are developed together with the post-processing tools required for the data analysis. The computational tools existing today are, mostly, private codes by the roller bearing manufacturers that are not available to the general community and that involve methodologies that are proprietary to the bearing manufacturers. This situation causes two fundamental problems: the codes cannot be used by third parties and the scientific competence cannot be validated by independent entities. Therefore, in the framework of the MAXBE project a code to overcome these situations started to be developed. While working for project MAXBE, the final goal of creating a computational tool that allows obtaining the full dynamic performance of a spherical or tapered roller bearing in working railway conditions was partially achieved. The theoretical models necessary to conclude the program are available. In order to achieve the final state of the dynamic analysis computational tool, extra time and effort should be spent in its creation. To do so, the BearDyn program should be submitted to some more testing and development, as well as code optimization to reduce the computational effort and allow results to be obtained more effectively. Optimizing the code should first focus in computation efficiency, eventually use parallel computational strategies. The other aspect that needs to be explored is the search for numerical methods well-adjusted to the type of mechanical system. The first approach to code optimization should be on code parallelization. By allowing parallel computing, the MATLAB® code runs in parallel in several processors of the computer, which can improve the code performance by allowing executions to run faster. Notice that at each time step, for the type of roller bearings used in railway axle boxes, these are about 1000 contact detection problems to be solved simultaneously, and independently. Another option that should be considered to improve the code is creating a numerical integrator that handles the particular aspects of the roller bearing dynamics. By doing so, it is possible to control every aspect of the integrator and optimize its performance to avoid problems similar to those identified in this study related to the integrator. In order to reduce the time spent in contact detection, also more accurate initial values can be calculated and delivered to the search method at each time step. While testing the implementation of contact detection in BearDyn, some cases were identified where the closest points were detected in opposite sides of the bearing, due to its symmetric geometry. This problem was solved by using the last correct results when this occurred, but it is not guaranteed to be the best solution. In order to avoid these problems, a more accurate way of predicting the location of two closest points at each time step can be developed. This can be done by, for
Copyright © MAXBE Consortium
9 of 33
example, predicting the location of the contact point on each time step by foreseeing the trajectory of each body instead of using the last successful values as initial guesses. In order to allow the monitoring of the bearings performance in railway operating conditions, as the final goal desired for project MAXBE, the dynamic response of the bearing should be converted to a vibration response using post-processing tools in order to be used as a basis for comparison with the dynamic response of bearings with defects. A final approach requires typical defects on bearings to be modelled in the code of the dynamic analysis tool. When this is achieved, the main goal of this project is reached, where the collection of vibration response data obtained with BearDyn for different bearings with or without defects can be used to infer the health of axle bearings, by comparison with the responses of bearings obtained via wayside or on-board monitoring systems. Validation of the models created by performing experimental tests is required to ensure the validity of the methodological assumptions and the consistency of the roller bearing models. Due to the impossibility for the roller bearing manufacturers to supply the detailed geometry of the bearing mechanical components, a thorough experimental campaign is required to properly verify internal geometries of roller bearings, their materials and lubricants in order to allow for the construction of meaningful models. It is estimated that the necessary resources to develop a suitable computational tool for the thorough dynamic analysis of roller bearings in railway operational conditions, including the experimental validation work and the integration of the models in the full vehicle model for realistic operation conditions, is about 36-48 person months.
Copyright © MAXBE Consortium
10 of 33
3. WP3 – ON-BOARD CONDITION MONITORING SYSTEMS 3.1. Wireless On-board monitoring system – SKF The rail transportation industry has seen a shift away from traditional time-based maintenance and toward condition-based maintenance. This shift in maintenance strategy requires the collection of critical data (such as temperature and vibration) from mechanical components to facilitate assessment of their condition. On-board condition monitoring systems offer the distinct advantage of allowing collection of this data during a journey and under different operating conditions. The prototype SKF Insight ™ on-board monitoring system developed as part of the MAXBE project is a condition monitoring system designed specifically for monitoring axle-bearings in passenger rail applications, although the technology in other embodiments is transferable to freight rail applications. These wireless, self-powered sensors are installed inside the axlebox, on the bearing outer seal. The sensor nodes acquire vibration, acoustic emission, temperature and speed measurements, and send them wirelessly to a gateway which routes the data to an on-board PC. This data is then sent to a remote database where it is monitored by SKF @ptitude Observer software.
Figure 3.1 - System Architecture
Copyright © MAXBE Consortium
11 of 33
Trials of various iterations of the prototype SKF Insight™ system have been performed with EMEF (2013, 2014) and later with Nomad Tech (2015) on an operational CP Alfa Pendular. The objective of these trials was to demonstrate the prototype sensors in a field application. Each trial lasted several months and allowed for the collection of a significant corpus of data. These trials facilitated stepwise improvements in sensor performance and provided successful demonstrations of sensing capabilities, wireless communication technology and integration of the data into an industrial condition monitoring software system (SKF @ptitude Observer). Four axleboxes were instrumented in each trial with SKF Insight™ prototype wireless sensors. Data from these four sensors were aggregated by a carriage-mounted wireless gateway which in turn was connected to an industrial PC located on board the carriage.
Figure 3.2 - Sensor mounted on a bearing seal (left) and visible during installation on the CP Alfa Pendular (right).
Figure 3.3 - View of the sensor installation once completed.
Wireless communication between sensor and gateway proved to be the greatest challenge initially. The environment at axle and bogie level is challenging from a wireless communications perspective with metal in and around the sensor location. This challenge was further compounded by the requirement for low power operation of the sensors. By locating the sensor antenna externally on the axle box it was possible to circumvent some of the issues with wireless communication, however achieving the optimal balance between network stability and low power operation was achieved in a stepwise manner through software and firmware modifications. Each successive trial was accompanied with marked improvements in performance such that the current ongoing trial with Nomad Tech has had 100% network reliability. Wireless communication technology when used in conjunction with measurement
Copyright © MAXBE Consortium
12 of 33
system facilitates the monitoring of machinery that was previously difficult, impractical or inaccessible and this trial has been a successful demonstration of this. Close cooperation with EMEF and Nomad Tech allowed the use-cases for the system to be documented and highlighted key areas for improvement including sensor design and data integration with their systems. The specific measurements employed in this system development were selected because of their proven capabilities in detecting bearing damage and damaging bearing conditions at an early stage. These measurements include SKF Acceleration Enveloping Band 3 (vibration), SKF Acoustic Emission Enveloping (acoustic emission) and temperature. The probability of damage detection in the trials, given only four wheels were instrumented, was low. Despite this, however, bearing activity was visible in the data albeit intermittently and at levels that did not indicate damage was present. Given the optimal location of the sensors in relation to the axle bearings, coupled with extensive experience in rail applications, it is expected that bearing damage or damaging activity (if present) would be clearly detectable. It is recommended that this installation location be maintained in further development due to its proximity to the axle bearing.
Figure 3.4 - Example SKF Acceleration Enveloping Band 3 spectrum (top) from the ongoing trial on a CP Alfa Pendular train in conjunction with Nomad Tech.
The benefit of the SKF Insight™ prototype system however is not limited to simple bearing damage detection but can be extended to also include maintenance strategy – with the potential for extending intervals between maintenance and also ensuring maintenance is performed on the right components at the right time. The bearing manufacturer (SKF in this instance) is central to informing this shift in maintenance strategy, and by providing a bearing-oriented measurement system (in addition to bearings) would be in an ideal position for providing real benefit to the customer. It is recommended that to ensure maximum benefit to the train operator, manufacturer or maintenance provider that the interconnection between condition monitoring system and bearing manufacturer be maintained in such system development for a rail application.
Copyright © MAXBE Consortium
13 of 33
3.2. On-board monitoring system developed by UPORTO and EVOLEO The on-board monitoring systems have the ability to monitor the asset more continuously, when compared with the wayside systems since the last ones are only able to monitor and diagnose when the train passes over the system. The continuously monitoring is a very important characteristic to understand and to interpret the evolution of the bearing failure modes by analyzing the trend of signals and information in short periods of time. Another important advantage of the on-board monitoring systems is that since they are installed in the train, they can be adjusted and developed considering the specifications and the unique characteristics of a particular axle bearing of a certain type of train and bogie. This feature enables an optimal adjustment of the hardware and even the diagnosis techniques implemented in the software of the monitoring system. However, on-board monitoring systems have still significant limitations: the hardware configuration that has to be sufficient robust to endure the hard conditions of the railway field; the systems require power supply. Considering the innovations and the advantages but also the problems and constraints stated above, within the frame of MAXBE project an onboard wired system was developed by UPORTO and EVOLEO with the support of NOMADTech to monitor the condition of the axle bearings based on the vibration and temperature measurements. This system integrates a data processor unit and allows the assessment of the real condition of the axle bearings and it is prepared to be installed in one bogie of the Alfa-Pendular train. The system is composed by a set of sensors installed in the axle box, a GPS and an acquisition system connected with a data processor unit installed in the on-board CPU. This on-board monitoring system measures two aspects of the axle bearing behavior: the temperature and the vibration levels in three directions. A general overview of the on-board monitoring system configuration is presented in Figure 3.5. The acquisition system receives the raw data from the sensors and stores it in an on-board CPU. The data management, such as correlating the data from the sensors with the GPS information, and the data processing are also performed in this on-board unit. The communication module allows the access to the raw and processed data, enables the remote access to the system and monitoring the state of the equipment. Then the raw data and the processed data converted into key performance indicators (KPIs) are transferred to an on-shore server, which enables the access of the data in the data fusion algorithms and feeds the smart diagnostic tool and the condition-based maintenance model. Considering the benefits of the correlation of on-board and wayside measurements, a radiofrequency identification system (RFID) is installed in the Alfa-Pendular train crossing the Estarreja test site.
GPS
RFID
Raw Data Sensors (Vibration and Temperature)
Raw Data
KPIs
- Raw Data Storage - KPI Storage - Remote Access
Acquisition System and On-board CPU
GPS
Figure 3.5 - General overview of the on-board monitoring system in Alfa-Pendular train (CPA4000)
Copyright © MAXBE Consortium
14 of 33
A detailed overview of the layout and installation of the system in presented in the following figure.
Figure 3.6 - Installation of the on-board monitoring system in Alfa-Pendular train (CPA4000)
The system is composed by a Dual Link Wireless Communications: 4G Link with dual MIMO antenna and WiFi AcessPoint (Local Connection). In terms of power supply, the system has an input of 100 Vdc@3A and 300 W and it acquires data through a Data Acquisition System (DAQ) and a three Axle Accelerometer [X,Y,Z], IEPE with a maximum range of 250G and an acquisition rate of 10 ksps. The temperature is measured by an RTD sensor up to 200ºC and an acquisition rate of 1 sps. The GPS has an acquisition rate of 1 sps. The software interface of the data acquisition system is presented in Figure 3.7.
Copyright © MAXBE Consortium
15 of 33
Figure 3.7 – Software acquisition interface of the on-board monitoring system in Alfa-Pendular train (CPA4000)
3.3. On-board monitoring system installed by UoB Figure 3.8 shows the onboard setup installed in UK. AE data were collected using an AE sensor which was attached to the housing with magnetic hold-downs. The sensor should be located as close to the bearing as possible, with a good mechanical pathway between the bearing and the sensors to assist in the transmission of vibration. The sensors should be fixed securely to the casing to ensure good transmission of the high-frequency vibration. The data were recorded by the operator placing inside the wagon. Experimental tests in Long Marston were completed using a freight wagon going forward with 30 mph and backward 20 mph speed. The length of the track was a few hundred meters and the wagons consist of Timken bearing model 9959199100. The sampling rate for acoustic emission signals was 500 kHz.
Copyright © MAXBE Consortium
16 of 33
Accelerometer
Figure 3.8 - Onboard mounting of the AE sensors and onboard for the assessment of axle bearings using magnetic hold downs
Amplitude difference in raw data of the onboard acoustic emission test can be observed in Figure 3.9. This indicates that the higher defect size causes the greater AE signature. Crest Factor failed to separate the bearings with different defect sizes.
Amplitude / V
5 8mm roller defect 4mm roller defect 2mm roller defect Healthy 0
-5 0
2
4
6 Time / s
8
10
12
Figure 3.9 - Raw data from AE analysis of the onboard Long Marston test – different defect sizes
Copyright © MAXBE Consortium
17 of 33
(a)
2.4 2.2 2
0
0.5
1
1.5
2
2.5
3
3.5
0.025
0.02
4
0
0.5
1
Defect size / mm 36.5
2
2.5
3
3.5
4
3
3.5
4
140
36 35.5 35
1.5
Defect size / mm
Kurtosis / V
C rest factor / Arbitrary units
0.03
2.6
R MS / V
PK-P K / V
2.8
0
0.5
1
1.5
2
2.5
3
3.5
120 100 80
4
0
0.5
1
Defect size / mm
1.5
2
2.5
Defect size / mm
0.4
15
0.3
10 5 0
Crest Factor / Arbitrary unit
RMS / V
20
0
1
2
3
4
5
6
7
0.2 0.1 0
8
0
1
2
Defect size / mm 70 60 50 40 30
0
1
2
3
4
5
Defect size / mm
3
4
5
6
7
8
6
7
8
Defect size / mm Kurtosis / Arbitrary unit
PK-PK / V
(b)
6
7
8
600 400 200 0
0
1
2
3
4
5
Defect size / mm
Figure 3.10 - Overall value changes of the bearing with roller defect by changing the defect size Onboard Long marston (a) vibration and (b) AE measuement.
Copyright © MAXBE Consortium
18 of 33
4. WP4 – WAYSIDE CONDITION MONITORING SYSTEMS 4.1. Vibration monitoring system for freight, urban and long-distance trains The most used technique for monitoring the axle bearing condition and the wheelset is by measuring and controlling the in-service temperature of the rolling stock axle boxes through a device installed in the track in pre-determined points of the railway network. However, this method has several limitations since it is based on thermal radiation of bearings which is affected by several factors such as the ambient temperature and other external effects such as the material of the axle box, the surface finish, the design and the operating conditions. Also, hot boxes detectors do not fulfill the most recent and advanced requirements of the railway industry regarding the detection of defects in early stages of development, because when the temperature alarm is activated, it usually means that axle bearing is in imminent failure. Nevertheless, trend analysis to the temperature levels can always be performed and used to provide predictive maintenance information. In this context, the development of a vibration wayside system within the MAXBE project aims to develop a competitive wayside system able to detect wheelset defects in freight, urban and loan-distance trains, based on the measurement of deformations in the superstructure elements of the track and the evaluation of the static and dynamic wheel axle loads correlated with defects of the wheel. Furthermore, the system is able to identify the total number of axles and the geometry, speed, acceleration and direction of the train composition, based on a pre-defined database. In order to provide historical data, correlations and tendencies, a RFID system is also employed allowing to detect not only the type of carriage, but which specific carriage is being registered and monitored. The vibration condition monitoring system is a competitive, modular and robust track side system able to detect wheel impacts and defects, and therefore by correlating the information with complementary monitoring systems, decrease the number of false alarms related with axle bearing failures. Moreover, the system has the potential to reduce maintenance costs and to improve ride quality, all of significant importance to train operators. At the same time, a significant factor to be considered by the rolling stock companies is maintaining their vehicles in such a way that they can run for as long as possible, thereby maximizing their value. The developed vibration system was installed along an equivalent wheel perimeter length in a section of the Portuguese Railway Northern Line. The instrumented strain gages were placed over a total length of 3.6 m that considers seven sleepers equally spaced in 0.6 m interval. The sensors are protected against the railway adverse environment, with dust, ballast, water and all the heavy maintenance activities performed in this type of infrastructure with a robust mechanical system. The cables connecting the sensors to the acquisition system are underground cable ducts in order to ensure the safety and durability of the system after the maintenance activities. The system configuration allows to weight in motion and to detect wheel defects in a speed range between 5 to 250 km/h, a wheel diameter between 350 and 1000 mm, and it is able to detect and to identify trains up to 300 axles with a range of load between 5 and 400 kN.
Copyright © MAXBE Consortium
19 of 33
Figure 4.1 – Vibration wayside monitoring system:
installation The developed monitoring systems are subsequently integrated into the MAXBE Monitoring System Integration platform, which is a global monitoring system that aggregates all the information generated by both the on-board and the wayside systems, such as the key performance indicators that results from the post processed data. The required general information that is provided by the infrastructure managers or by the rolling stock operators may also be included in the integrated system. Although the platform is developed based on the requirements of the project within the scope of WP5, it will be able to integrate different monitoring systems that can be developed in the future or that are already in the market, such as the hot box detector systems. The tool is able to simplify and to summarize the information gathered by the several monitoring systems and will be used by the infrastructure managers, by the rolling stock operators or by the maintenance managers as an aggregating tool correlating the data from the monitoring systems in order to get a more accurate and reliable monitoring and diagnostic tool that supports the decisions of the responsible stakeholders.
Copyright © MAXBE Consortium
20 of 33
4.2. LRV wayside monitoring Light rail has very specific needs, both technically as economically. In depots and workshops, mainline monitoring systems are being deployed, but their cost is often too high for a small network. Their use outside the depot for in-service monitoring is often impossible due to the accessibility to the public of the LRV track. For axle bearing maintenance, scheduled maintenance is the norm and blocked wheels due to bearing damage occur. State-of-the art monitoring systems based on vibration monitoring exist to detect wheel related faults such as wheel flats and oval wheels. They allow for an early alert to prevent further damage to the vehicle (notably the bearings) and the track. As a conclusion of the work in MAXBE, we can propose the extension of such existing systems with vibration sensors with a high frequency range and distributed over a length of track equal to the wheel circumference. The data acquisition and processing hardware remains the same and only the processing software needs to be extended. It is expected that this will result in an extra cost of 25% in the of the conventional LRV wayside monitoring system. The results obtained in MAXBE for LRVs showed a complementarity of this technique with early detection of axle bearing deterioration. A limitation of the track-based detection for resilient wheels is that the early onset of a developing axle bearing fault cannot be detected. However, the detection will allow detecting damage well before risk of wheel blocking. This explains the business case for this type in system as an addition of the standard wheel fault detection system. As further work, it can also be investigated whether transmission problems can be detected on the same principles. Further extensive large-scale testing will also be necessary to determine the optimal distribution of wayside monitoring systems to guarantee the effective monitoring of a complete fleet.
. Figure 4.2 - LRV wheel-out-of-roundness monitoring system in Antwerp
Copyright © MAXBE Consortium
21 of 33
4.3. Acoustic emission monitoring system Another customized wayside monitoring system is based on a set of AE sensors mounted on the rails, connected to a data acquisition unit and a data processor module installed trackside. The data were recorded after automatically triggering the system to acquire as the train neared the instrumented section of the track. Wheel and rail form a direct mechanical path for the acoustic emission signals produced from the bearing to be transmitted to the AE sensors mounted on the rail i.e. the detection zone. Detecting bearing defect signals transmitted via this direct mechanical path provides more accurate results compared to airborne acoustic detection, because it eliminates adverse effects of surrounding noises and other environmental parameters, such as wind and aerodynamic forces. Moreover, the signals acquired contain high frequency information which makes detection of faults more likely. Figure 4.3 shows the installation outline of the wayside tests carried out in Long Marston. The tank freight wagon containing the three faulty bearings was towed by a locomotive as shown in Figure 4.4. The faulty bearings were only at the same side of the second, third and forth wheelsets of the freight wagon with 2, 4 and 8 mm roller defects, respectively. Tests were carried out at up to a maximum speed of 48 km/h over a straight section of welded track approximately 1000 metres in length. The sampling rate for AE signals was 500 kHz and the duration of the acquisition was set at 12 seconds.
Figure 4.3 - Simplified outline of the wayside installation configuration.
Copyright © MAXBE Consortium
22 of 33
Figure 4.4 - The Long Marston testing configuration. The yellow locomotive pulling the test freight wagon can be seen in the back.
An optical unit capable of measuring the speed of the train and counting the number of wheelsets was employed in order to correlate AE signals with the position of the wheels. The optical unit was also used to trigger the data acquisition unit to acquire data while the train was passing through the detection zone. As the system was also counting the number of axle boxes, it is possible to truncate the AE signal exactly at the time that each wheelset passes through the detection zone. This method saves a vast amount of data as the sampling rate of 500 KHz needs a large volume of physical memory. This also makes the analysis period shorter. Figure 4.5 shows the instrumentation used for both laboratory and field tests in this work. Amplicon desktop computer Data acquisition board
4 channel hub
Amplifier Figure 4.5 - Part of the wayside data collection system.
Moving RMS and moving kurtosis are time domain signal processing methods, which are applied to the AE data. The results are shown in Figure 4.6. In the moving RMS bearing defects are clearly evident with higher amplitude compared to the engine noise although the amplitude of the unknown noise is still high. The amplitude of the unknown noise is reduced in the moving Copyright © MAXBE Consortium
23 of 33
kurtosis analysis. However, in this case the amplitude of the engine noise is increased. The bearing defects are detectable in both of these analysis methods, therefore combining the moving RMS and kurtosis can improve the results. 10
Amplitude / V
8 6
7.43 sec
4
6.38 sec
2 0 0
(a)
2
4
5.45 sec
6 Time / s
8
10
12
Kurtosis / Arbitrary unit
80
60
6.51 sec
40
20
5.57 sec
3.03 sec
(b)
0 0
2
4
6 Time / s
8
10
12
Figure 4.6 - Moving RMS (a) and moving kurtosis (b) diagrams of wayside AE measurement.
The existence of high amplitude AE waves in the raw data can reduce the capability of fault detection. Frequency domain analysis can be used to overcome this problem. Figure 4.7 is the spectrogram (time-frequency diagram) of the measurement and shows that the unknown noise remains dominant, this can falsely identify a defect.
Figure 4.7 - Spectrogram of AE data from wayside field test
Figure 4.8 demonstrated the TSK analysis. High amplitude peaks from the bearing defect do not indicate a normal distribution behaviour hence the value of kurtosis rises up to 260 in certain frequency bands. However, the unknown noise contains a lower value of kurtosis in all frequency bands. TSK of bearing defects has much higher value in high frequency components compared to the unknown noise. Therefore the TSK is capable of distiguishing between these
Copyright © MAXBE Consortium
24 of 33
two peaks. Engine noises contain lower amplitude in high frequency ranges. Therefore using a high-pass filter (fc = 150 kHz) before the TSK analysis can remove the engine noise from the data. The high amplitude peaks in raw data which have higher TSK value in frequencies between 150-250 kHz are related to the bearing defects. In addition engine noise usually has lower amplitude compared to the bearing defect peaks in raw data. Using a threshold-based method can also be used to eliminate the engine noise.
100
Bearing defect
Engine noise
Brake noise.
K urtosis/A rbitrary unit
80
60
400 25040 200
200 150 0 0
20
10 2
4
50 6
8
10
12
0
0
Frequency / kHz
Time / s
Figure 4.8 - Time spectral Kurtosis diagram of AE data from wayside field test.
Figure 4.9 shows the wheel flat induced on one of the wheels for testing purposes.
Figure 4.9 - Photograph of the wheel flat induced for testing purposes.
Copyright © MAXBE Consortium
25 of 33
5. WP5 – INTEGRATION OF WAYSIDE AND ONBOARD MONITORING SYSTEMS 5.1. Wayside and on-board integration software Nowadays, many monitoring systems to improve the safety and reliability levels of the railway infrastructure and rolling stock are available on the market; for this reason, due to the rich variety of sensors and devices (e.g. type, technology, etc.), the main MAXBE-WP5 objective was to develop a suitable and reliable system to integrate such monitoring solutions (whether they are installed wayside or on board), and to extract the most useful information of the functional status of the railway assets from the fusion of the heterogeneous and asynchronous gathered data. Basically, the integration of wayside and on-board monitoring systems is a crucial challenge aiming at correlating and aligning measurements and data, coming from different heterogeneous sources, and related to the functional behavior of a particular asset, which must be univocally identified. This task must be accurately performed, on the one hand, to extract a rich picture of the monitored asset status, and, on the other hand, to avoid loss of important diagnostic information.
Figure 5.1 - Three-levels architecture
Based on the experience gained in the context of the MAXBE project, an Integration Software Platform should be able to collect, store, process and visualize the measurements and data coming from the different monitoring and inspection technologies. The system should aim at providing to the maintenance operators synthetic and synchronized diagnostic information (such as alarms, KPIs, etc.) about the functional condition of the railway assets under examination, and supporting them in the decision-making processes. In the context of systems integration, an overall architecture able to support the desired functionalities of the Integration Software Platform should be defined; concerning the MAXBE project, the considered architecture has been divided into three main levels (see FIGURE 5.1), which will be described below based on a “bottom-up” approach.
Copyright © MAXBE Consortium
26 of 33
The “lower” level, called Monitoring Systems Level, represents all the monitoring systems composed by different kinds of sensors that can be installed along a railway line and on-board train, which are able to collect on-field measurements and data related to the functional status of one or more assets under examination. In order to guarantee the interoperability of the different monitoring systems, a common format for the representation of data should be defined. From a technical point of view, this format should represent the common interface between all the different types of sensors and devices installed onto the train and along the railway line. For instance, the eXtensible Markup Language (XML) represents one of the possible suitable solutions to standardize and normalize the collected data since it allows defining a common data file format. Indeed, a set of custom XML file formats has been defined during the MAXBE project in order to satisfy the fundamental requirement of interoperability among heterogeneous monitoring systems. The choice for a custom XML file format, during the Common Interfaces definition phase of this project, has been guided by the lack of a common standard format for information exchange. However, new standard formats should be used in railway applications such as the one defined by the RailML Initiative, which is currently used in the context of many railway on-going research activities by both academic institutions and companies. The “medium” level, named the Storage and Data Management Level, is responsible for the storage and management of data coming from the lower level of the architecture. Usually, this layer includes several software systems (such as database management systems, file servers, network storage solutions, etc.) with the aim of storing data and making them available to all the applications interested in their processing (e.g. for condition monitoring applications). In the MAXBE project, it has been decided to store XML data files into a central dedicated FTP server, since it represented the best storage solution for the purposes of the project due to its simplicity of usage. However, following the current research trends in condition monitoring and diagnostics, it becomes more and more important to build storage infrastructures able to deal with large amounts of data, for instance by exploiting state of the art Big Data technologies, which would also give the possibility of performing powerful statistical and data mining analysis over the collected historical data. The “top” level, called the Application Level, includes all the software technologies developed for data processing and visualization purposes. For instance, in the MAXBE project architecture, this layer is represented by the Data Fusion Tool (DFT) and the Human-Machine Interface (HMI). The DFT performs the integration of all the data coming from the two lower levels, and extracts synthetic diagnostic information related to the functional status of the railway asset under examination. The HMI, instead, shows data in order to support the final users (such as maintenance operators, infrastructure managers, etc.) on the decision-making processes and on the optimization of the maintenance activities planning. The DFT should be designed to be modular, scalable, flexible and configurable in order to be interfaced with multiple monitoring systems; it should be able to generate high-level and highinformative patterns from raw data that include all the possible information on the functional status of the axle bearings. The extracted information should be easily accessible to the final users (such as maintenance operators, infrastructure managers, etc.), and the system should highlight critical events and conditions. In order to develop a Data Fusion Tool architecture, one of the potential solution that can be taken into account is the Joint Directors of Laboratories (JDL) model. This model generally shows all the possible functions that a Data Fusion Tool could include and standardizes the lexicon about those functionalities. The Data Fusion Tool users should carry out its functions by using a Human Machine Interface (HMI): in the context of the MAXBE project, it is constituted by a web-based visualization
Copyright © MAXBE Consortium
27 of 33
system that allows visualizing data and information generated by the DFT, such as potential alarms, in a user-friendly graphical way. The most important characteristics of such an interface should be the clearance, the simplicity and the ease of use. Moreover, it should include pop-out elements that highlight the presence of dangerous situations, and consequently it should be able to explore data related to the particular phenomena in order to have, if requested, a more precise view of the situation under examination. The developed HMI for MAXBE is composed by a “Main Panel” that visualizes in real time all the information related to a train transit and allows managing alarms. The details of events (e.g. vehicles, axles, wheels, axle bearings, etc.) and alarms are shown into new windows or pop-ups (see Figure 5.2 ) that can be managed by the operators to support the maintenance activities planning.
Figure 5.2 - Example of HMI: “Transit Detail” pop-up (in foreground) and the “Main Panel” (in background)
As a final remark, the architecture of the system developed for the MAXBE project is general and could be successfully used not only for axle bearing, but also for supporting condition monitoring of any railway asset. Moreover, in the future, new research activities could be performed in order to implement innovative functionalities into the system; for instance, the availability of historical data should allow supporting the development of predictive models (as done into the WP7 of this project), which are able to study and predict the asset functional behaviour and the related degradation processes for maintenance decision-making purposes.
5.2. Tool for optimal physical distribution of the monitoring systems In a context where financial pressure on Infrastructure Managers is increasing, prioritizing investments in an effective way is of paramount importance. For this reason, within the MAXBE project a software tool has been developed in order to support Infrastructure Managers in deciding the optimal distribution layout of the monitoring systems. By doing this, Infrastructure Managers are able to make planning of diagnostic systems more effective and provide usability for safety, operational and vehicle maintenance reasons.
Copyright © MAXBE Consortium
28 of 33
The criteria of prioritization plays a fundamental role in the software and for that reason, it has been defined according to the extensive know-how of MAXBE partners, with the noteworthy contribution of REFER (Portuguese Infrastructure Manager). The main parameters for prioritizing chosen were: speed, traffic volume, type of train, accidentability rate and distance between existing monitoring systems. The software tool has been designed as to be valid for any EU country and hence, it is flexible to account for differences in regulations, vision and needs between countries. In this sense, main regulations for several EU countries are already included in the software (i.e. Spain, Portugal, Germany, Italy, France and United Kingdom). The outcome of the developed tool are suggestions for the most fitting places for track-side diagnostic systems inside of a respective network. The SW tool offers the following advantages: • • • • •
Optimisation of planning process Optimization of distribution of HBD/TBD can save costs Planning in terms of country-dependent installation requirements Traffic, infrastructure and safety dependent parameters can be covered One planning standard for European countries
For further development it can be recommended to integrate more parameters for an optimal physical distribution of the wayside diagnostic systems and to consider a more detailed decision matrix based on the experiences received from potential users. Another important issue is to aim for an opportunity to exchange data between the developed SW tool with already established asset management systems like RailSys®. With that it would be possible to transfer data between the infrastructure databases of railways (e. g. network and line data, gradient, signal or switch positions, allowed speed) and the here developed tool more easily. The tool allows deciding on the length of the segments into which the network is divided, thus enabling to use the same discretization of the network considered in the asset management system. However, the transfer of data should further studied. For further development beyond this project the SW tool should aim to interact with asset management tools, to offer the following advantages: • • • •
Choosing track sections from infrastructure model with real asset data Also respecting real operational data (number and type of trains) New SW tool can use data from asset management model Detailed planning of wayside device locations for existing or future networks
Copyright © MAXBE Consortium
29 of 33
6. WP7 – ASSET MANAGEMENT TECHNOLOGIES 6.1. Condition-based Maintenance Schedule model In Europe the so-called ECM (Entity in charge of maintenance) is responsible for rolling stock maintenance, regulated by a new EU-Norm. That ECM can be responsible for freight trains (actual rule) or passenger trains (not now, but maybe in the future) operating company, for example. This entity plans and conducts maintenance for the train fleets of their customers. Usually these customers are the owners of the vehicles, e.g. TOC or traffic organising companies or authorities. Based on additional information from COMSA any company who wants to be an ECM has to obtain a certificate which proves that they are able to maintain accordingly the freight wagons. The ECM may outsource some of their tasks, but it is always responsible for the maintenance of the wagons. The ECM is receiving the required maintenance intervals for the vehicles from the vehicle owners. It is the ECM's own interest to optimize the maintenance they are responsible for, with the aim to meet these required intervals at least. One important issue is the requirement of strictly classifying the type of rolling stock that have to be handled in maintenance operation. Usually the maintenance for locomotives and passenger wagons is done based on time and kilometric performance. For freight wagons only the time since the last stay in the depot is relevant for planning maintenance in Germany whereas in Spain also distance travelled can be an indicator. The optimizer developed by UCC and partners can handle time alone, kilometric alone or both metrics. The optimiser allows to plan and value the effort of the maintenance strategy. This is done with a comparison of the costs assigning resources for conducting condition-based maintenance, such like personnel, equipment and deport capacity, with the cost savings achieved using a condition-based maintenance strategy instead of a conventional strategy. That decision also includes the question, if a combination with a corrective with a preventive exam is useful and possible. The effort and induced costs for taking the trains out of productive service and to and from the depot is also relevant. At least the ECM has to guarantee for the availability of trains written down in traffic contracts of the train operating companies. Furthermore the developed optimiser illustrates why condition-based maintenance of certain components, such as axle bearings, should not be considered in isolation when implementing a fully optimal approach. There are components that are not monitored that require maintenance, which will be performed, as now, following manufacturer specifications for periodicity of maintenance (be it based on time or kilometric performance). Therefore one must consider, and indeed schedule, both condition-based and periodic maintenance in a centralised system, since depot resources and staff are shared amongst all maintenance works. It was also noted that the uncertainty in the scheduling problem is caused not only by conditionbased maintenance of components such as axle bearings, but also by variation in the maintenance duration and resource requirements of planned maintenance actions. The (open source) SW included in the maintenance optimiser contains a tool for automatically modelling the distribution of maintenance durations, or staff/resource requirements for a maintenance task based on historical maintenance data, enabling the maintenance operator to adapt the expected duration and resources required for a maintenance task over time. The maintenance optimiser tool also contains proactive and reactive techniques for handling the uncertainty inherent in the maintenance-scheduling problem. The developed optimiser contains functionality to cover actual and future requirements of the railway industry regarding condition-based maintenance for vehicles and components.
Copyright © MAXBE Consortium
30 of 33
The results of the work undertaken in the design, development and evaluation of a maintenance schedule optimisation tool can be found in deliverable documentation and software prototypes of WP7.
6.2. Smart Diagnostic 6.2.1. Smart diagnostics
Research and development of smart diagnostics in WP7 has demonstrated that the training and implementation of Artificial Neural Networks (ANN) can bring enhanced diagnostic capability to traditional rule-based approaches. It is therefore recommended that smart diagnostic techniques are developed further for future implementation in the monitoring of axle bearings and other train mounted components and sub-systems. Moreover, as the dependency on monitoring systems increases for the planning of on-condition maintenance, it becomes ever more important to generate reliable and accurate fault diagnostics. Asset monitoring systems must serve as an improvement to safety and business performance rather than a hindrance, and therefore must be robust in their diagnostic output and decisionmakers must have confidence in the information generated. Unreliable diagnostics, such as imprecise or incorrect fault (and 'no fault') reporting could lead to maintenance tasks being under-performed or over-performed, with the resulting reduction in safety and reliability, or increase in unnecessary maintenance costs respectively. 6.2.2. Information Integration
A maintenance strategy associated with condition-based maintenance and predictive maintenance relies on maintenance engineers having full knowledge of the condition of their rolling stock assets at any instance in time, as well understanding how asset condition will change in the future. In order to assist in decision-making regarding asset management and maintenance, it is recommended that as much relevant information as possible is easily and directly available to engineering management staff. On one hand it is important that asset condition information generated by different monitoring sources is contained and visually displayed to the engineer in the same place. For wheelsets, this could include information from systems such as axle bearing monitoring, wheel profile measurement, wheel flat detection, static load distribution, vibration monitoring and diameter monitoring. On the other hand, most maintenance organisations in the EU are ran as business entities, and therefore should continuously work to improve key business performance indicators. A direct visualisation of asset condition together with business performance figures through a dashboard-style approach is a proposed method of improving core understanding of the functions within the business, and the links between fleet technical performance and organisational business performance.
Copyright © MAXBE Consortium
31 of 33
7. RECOMMENDATIONS Through the maxbe results and the consultation with potential end users, a number of recommendations can be made: • •
• •
•
Axle bearing monitoring should be seen as an extension to Wheel Flat and Out-ofRoundness Monitoring systems. The monitoring of the axle bearing requires the knowledge of the frequency response of the healthy bearings as a necessary data for the understanding of any abnormal condition. To obtain this knowledge extensive experimental tests in axle bearings are needed as well as defining proven computational tools for the assessment of reliable and fundamental dynamic response of the axle bearings in realistic operation conditions. More accurate methods (than measuring and controlling the in-service temperature of rolling stock axle boxes), such as the vibration, the acoustic and the grease analysis, should be brought to the design of axle bearing monitoring systems. The monitoring of axle bearings should consider the possibility of adopting in some cases onboard condition monitoring systems that offer the advantage of collecting measurements during a train journey and under different operating conditions. Also the combination of wayside with onboard systems with the integration of the measurements from both systems may be advantageous in some situations. Due to the limited scientific and technological knowledge available in the railway industry regarding the onboard monitoring systems and also considering the lack of experience in the development, installation, maintenance and knowledge extraction of on-board systems, there is a need to contribute to the development of the Technical Specifications for Interoperability of the rolling stock and to the requirements and specifications of on-board equipment for axle bearing condition monitoring systems. Also, the requirements regarding the monitoring performance, trends, thresholds, and limits for alarm levels, operation requirements and interface with other systems have to be clearly established for this novel monitoring systems.
•
The MAXBE partners have developed and tested a data integration architecture integrating all state-of-the-art and newly developed axle bearing condition monitoring techniques, both for on-board and for wayside monitoring. The architecture contains a specification for the data representation layer, detailing data representation for data on DUT (device under test), i.e. vehicles and axle box and for the test equipment, i.e. the sensors and the monitoring system (data acquisition and processing). The architecture also specifies the protocol to be used at the data communication layer. The standard specification is available as MAXBE deliverable 5.1. To realise the vision of interoperable maintenance and an open market for condition-based maintenance, we propose adding the above standard for interoperable axle bearing condition monitoring data to the Rolling Stock TSI. By adding this specification to the Rolling Stock TSI, an uptake for the LRV market and for wayside systems can also be envisaged.
•
Tools for the optimal physical distribution of wayside monitoring systems should be adopted by railway administrations in order to: optimize the number of systems in a railway line without compromising the detection of axle bearing failures at an early stage; contribute for the standardization and the interoperability of the systems’ installation.
Copyright © MAXBE Consortium
32 of 33
•
A shift from traditional periodic-based maintenance to a condition-based and predictive maintenance approach can bring clear benefits to the maintenance organisation in terms of costs and efficiencies, and to the train operator in terms of improved reliability, increased fleet availability and most importantly improved safety. However, it is not at all trivial to effectively schedule condition-based maintenance tasks (which could be related to any train sub-system) alongside planned preventive maintenance tasks for all train units in a fleet or indeed across multiple fleets. It is therefore recommended that maintenance and operator organisations make use of planning and optimisation tools such as the one developed in MAXBE to assist with the task of maintenance scheduling and to optimise key objectives within the organisation. This way, tasks can be scheduled automatically and updated as soon as there is a perturbation to the planned schedule. Without an 'optimiser' tool, planning decisions are harder to make and resources may not be managed in the most effective and efficient manner. The need for computationally-assisted planning will increase as organisations rely more heavily on automated asset monitoring systems (such as axle bearing monitoring) and more maintenance tasks are planned based on monitored asset condition, rather than pre-determined frequencies for inspection and maintenance.
Copyright © MAXBE Consortium
33 of 33
