Predictive Maintenance Team in Assystem Digital Factory Division working on development of smart condition monitoring solutions to support manufacturing companies to improve their operational efficiency, reduce costs and maximize productivity by turning operational data into actionable insights. The Predictive Maintenance techniques can predict asset failure times, detect anomalies (i.e. incipient faults) by accessing multiple data sources in real time so that the clients/organizations can avoid costly system downtime and improve reliability, availability and safety in their organization. The early identification/detection of asset failures can help maintenance technicians to minimize system downtime, deploy limited maintenance resources more cost-effectively and enhance flexibility and efficiency of their systems.
Assystem Digital Factory aiming to offer a real-time smart condition monitoring services so called "Digital Twins" for its clients to improve reliability, availability and safety of their equipment via smart predictive maintenance solutions. Hence, in this context, efficient diagnostics and prognostics techniques will be developed under "Digital Twins" program for Railway Transportation Asset (i.e. bogies, rails, wheels, point machines, etc.) Condition Monitoring.
Different mechanical faults in Railway infrastructure will be the main course of this research by using different analytical methods such as Electrical Signature Analysis (ESA). At first adequate sensors and data acquisition systems should be studied on the basis of a defined selection method, in which is taken into account the required measurement data, external influences, sampling rate, etc. For the test(s)/data collection a test protocol will have to be written in advance.
Different testing regimes will be utilized, starting from very simple single faults up to mimicking the complex multiple faults than can occur on the Railway systems. The electrical data will be collected together with data from the accelerometers and other alternative sensors.
The electrical signals will be processed to obtain the frequency spectrum and then faults would have to be identified and the failure time has to be predicted by the developed approach. The result of this study has to be summarized by the student as a report towards an application of predictive maintenance in a proof of concept "Digital Twins" for Railway transportation.
Skills and Knowledge
- Studying for Bachelor or Master degree in Engineering
- Strong research and analytical skills
- Strong knowledge in Signal processing techniques.
- Knowledge in fault diagnostics, prognostics and condition monitoring is a plus
- Programming skills: Python
- Knowledge in Machine Learning, Deep Learning and Statistical Analysis
- Fluency in English
Duration : The expected duration is 6 month at least. He/She will be preferably work 5 days a week from our site in La Defense/Courbevoie