Examensarbete: Analysis of bearing fault detector data

Analysis of bearing fault detector data Background Bearing faults can cause severe railway accidents, including derailments that pose a serious risk to human life. Trafikverket has installed trackside detectors that measure the infrared radiation to estiamte the temperature of the axle box. However, radiation from the wheel tread can also reach the axle box temperature sensor and affect its readings, potentially triggering false alarms. By developing a filtering method based on available data, the rate of false positives can be reduced while maintaining the rate for true positives. Relevant data include left and right axle box temperature, wheel thread temperature, ambient temperature, bearing make and model, snow/ice. Trafikverket has also installed an acoustic bearing fault detector. Its effectiveness could be evaluated by analysing measurements such as peak sound levels, RMS sound levels, and frequency-band data. Problem description, tasks and goals Develop a filtering method using available data to improve alert thresholds. Compare the effectiveness of infrared and acoustic detectors in identifying bearing faults. Category to which this degree project corresponds the best: Tech, Signalling, Electric, Data analysis. Type of degree project (can be both) ☒ Master (20 weeks) ☒ Bachelor/Högskoleingenjör (10 weeks) Language for the thesis ☒ Swedish and/or ☒English Is Swedish a language requirement? ☒ No, but Swedish is a requirement for future employment. Possibility to work from our office ☒ Yes ☐ No Address: Hälsingegatan 49, 113 31 Stockholm Contact person Mattias Wang Role: Vehicle/Maintenance Engineer Please contact me if you have any questions: mattias.wang@greencargo.com