The problem can be solved using a simple image classifier which predicts 1 or 0 depending whether its normal or faulty. I don't think it a time series problem (it is for a very short duration in the sense that , if it is faulty now, it will be faulty for some time ). Hence this can be solved using a normal image classifier .
I have completed several certifications on Deep Learning and machine learning from Coursera.
I have worked on Predictive Analytics,data mining, Object Detection and Object Recognition problems like counting the different number of vehicle type like cars , scooters, buses etc. I am also a freelance chatbot developer and have worked on NLP, NLU and RNN for these.
Some of the skills I possess are -
1) Python
2) Javascript
3) Machine Learning - Deep Learning, Sentiment Analysis , RNN
4) Chatbot integrations for Skype, Skype for business, Skype, Facebook
5) Backend - AWS
Let me know if you need anything else.