Digital railway system - collecting and utilising BigData
DLR
Collecting large amounts of data to derive insights is the great added value of increasing digitalisation. The railway system can also benefit from big data. To this end, DLR is developing processes and tools ranging from data acquisition (sensor technology), fusion and enrichment to data analysis and utilisation. The central tools for data acquisition are low-cost off-the-shelf solutions that can be used to collect data quickly, cheaply and flexibly on regular trains.
For monitoring the rail network, this approach offers an alternative to the time-consuming and expensive use of measuring trains. At the same time, it means an enormous increase in the quantity of data when data from many regular train journeys across the entire rail network enriches the individual point measurements. If this data is analysed using intelligent algorithms, the maintenance and servicing of the rail infrastructure can take a quantum leap with the help of digitalisation if maintenance is carried out on a preventative basis instead of at fixed intervals. This reduces costs, makes personnel and spare parts planning more efficient and prevents breakdowns and therefore delays.
The key to this big data approach is that the recording technology fits into a small case. It contains a software platform that can record the data, analyse it in real time and communicate filtered information to a central data platform. Depending on requirements, a wide variety of sensors such as cameras, microphones, inertial sensors or GNSS-based localisation components can be connected. High-frequency data is collected in this way. For example, vibrations are measured, which can be used to recognise track faults. Or the noise level is recorded using an integrated microphone, which can provide a network-wide overview of noise emissions if the data is extensive. Viewed over time, this also allows conclusions to be drawn about damaged tracks. Measurements of energy consumption along the route can reveal optimisation potential for driver assistance systems. Permanent localisation of the vehicles enables a real-time view of the overall utilisation of the network and the direction of travel of the vehicles. This also lays the foundation for future traffic automation. The recorded data is transmitted via GSM to a server, which enables all data to be analysed over a longer period of time. A future forecast of changes to an infrastructure is now within reach thanks to long-term data collection. The versatile software is used on different systems. For example, it can send current position data from a smartphone. Or it can be integrated into construction site lamps to independently mark the start and end of a construction site in a digital railway system, which can be visualised as a danger zone for dispatchers and train drivers.