Transitioning the railway sector to condition-based maintenance: from raw data to optimal planning.

We have finalized our design with an industrial grade IOT modem, we have also added an IMU to our system. In the final design we save all raw data each 10 minutes on SDCard. Our data storage capacity has been increased from few megabyte that was initial plan by using NOR Flash memory chips to 64GB by using a high speed SDCard. We also save the GPS data that we get from the modem. So we tag each file of recording with the time and place of the acquisition and later data scientists can analyse data based on the location of the train and its speed. We also save the acceleration and inclination of the train in 3D in the header of each file. This gives the experts possibility to normalize their vibration analysis with the mechanical loads that has been applied on bogies while each asquisition. We have added the possibility to our system to configure the input boards with different type of sensors. Although the accelerometers that we have used are induestrial sensors with IEPE standard, but it is possible that in some cases the data analysing team needs another type of sensors with specific interface. So, we have considered universal voltage interface and also temperature sensors in our design. In the firmware, we finally decided to make a boot loader to start the firmware from SDCard. So, if the maintenance team wants to collect data for longer time periode on one train even with shorter samplling periodes, they can easily do it now by making two copies of SDCards and replacing them when the system is close to fill the current SDCard. We have tested this mechanism for more than two weeks and it worked well. So, it is much more than enough for train trips in Norway that are takes few days. We have aslo devlopped aconsule for debugging the system and communication with the system while it is running the program. So, the maintenance team can be sure about the performance of the system while it is logging the data and is mounted on a train.

Leveraging the power of technology, we have successfully developed a system to relentlessly monitor the health condition of train bogies while they are operational on the tracks. This innovative approach revolutionizes our ability to maintain the safety and efficiency of our railway systems. The beauty of this system lies in its potential for precision. The maintenance team, traditionally reliant on predetermined schedules based on estimated running hours, can now utilize a more accurate forecast for maintenance activities. This is achieved by focusing on the conditions of each specific bogie, as revealed by the ongoing monitoring technology. The ability to customize maintenance schedules for individual bogies dramatically enhances the accuracy and effectiveness of our maintenance operations. A significant advantage of this development is its potential for reducing unnecessary overhauls. Traditionally, these overhauls are scheduled based on a one-size-fits-all estimation, often leading to wasteful and environmentally detrimental activities. By tailoring maintenance to the needs of each bogie, we can greatly minimize these unnecessary overhauls, resulting in substantial cost savings and a reduced environmental footprint. Moreover, adopting condition monitoring and predictive maintenance strategies adds another layer of safety to our systems. By continuously evaluating the health of the bogies, our technology can immediately signal the need for an overhaul at the earliest signs of degradation. This early warning system ensures that necessary maintenance is never delayed, thereby significantly increasing the safety of our trains. In essence, our trains are safer, more efficient, and more environmentally friendly as they traverse the tracks with our technology monitoring their every move. We have effectively transformed the traditional maintenance approach into a more precise, cost-effective, and eco-friendly one. Furthermore, by ensuring a higher level of safety, we boost passenger confidence in our railway systems, contributing to a more reliable and sustainable transportation infrastructure for the future.

The railway sector is one of the foundations of the country's mobility. Only in Europe, rail represents 45% of public transport, accounts for 26.9 billion passenger trips each and is used to carry 18.3% of the freight goods inland. One of the main costs drivers in this sector is the maintenance of the rolling stock, which is estimated to average between 24 and 34 NOK per unit per kilometer travelled. Achieving even a small reduction in these costs can translate into substantial savings, and have a crucial impact on the overall sector's performance. Traditionally, rolling stock preventive maintenance has been based on activities that are carried out on very conservative intervals. For that reason, many activities are performed more frequently than truly required. Resources are wasted on redundant check-ups, unnecessary replacement parts and the idle time during which the rolling stock is taken out of service. By properly monitoring the real-time condition of the rolling stock, however, it is often possible to predict the need for maintenance. With this kind of information, maintenance can be carried out only when strictly necessary. This new paradigm, called condition-based maintenance or CBM, poses multiple challenges: processing large amounts of data from multiple sources, building accurate models for failure prediction and re-designing the way maintenance providers carry out their operations. The main goal of this project is to lay the technical groundwork for the transition of the Norwegian railway sector to CBM. This will result in more efficient operations and lower maintenance costs. Mantena will become more competitive in the European railway market, and its customers will be able to provide a more frequent and reliable service to its passengers.