Sai kumar Chilukuri | SCSVMV UNIVERSITY, KANCHIPURAM (original) (raw)
Uploads
Papers by Sai kumar Chilukuri
As industry is getting advanced day by day incorporating new equipment's on a large scale, there ... more As industry is getting advanced day by day incorporating new equipment's on a large scale, there is a need to predict the machine lifetime in order to support the supply chain management. There are many ways of substitutions or upgradations required for any machine over a certain period of time, where its maintenance has become a major challenge. This problem is solved by building an effective predictive maintenance system which provides an intense spotlight for all types of machine industries. The log data is collected from the daily system activity from machines through deployment of various sensors facilitated to monitor the current state of equipment. A huge volume of numerical log data set is analyzed by the system for preparing the time series data for training and analyzing the model. Further steps involve bypassing the anomalies and fetching the clean data. The model is tested focusing on restoration time of any machine. This paper identifies and predicts failures of heavy machines, thus facilitating the predictive maintenance scenario for effective working of the machine at all situations. This work is implemented by LSTM network model for gaining authentic results with numeric data which facilitates major cost savings and offers higher maintenance predictability rate.
As industry is getting advanced day by day incorporating new equipment's on a large scale, there ... more As industry is getting advanced day by day incorporating new equipment's on a large scale, there is a need to predict the machine lifetime in order to support the supply chain management. There are many ways of substitutions or upgradations required for any machine over a certain period of time, where its maintenance has become a major challenge. This problem is solved by building an effective predictive maintenance system which provides an intense spotlight for all types of machine industries. The log data is collected from the daily system activity from machines through deployment of various sensors facilitated to monitor the current state of equipment. A huge volume of numerical log data set is analyzed by the system for preparing the time series data for training and analyzing the model. Further steps involve bypassing the anomalies and fetching the clean data. The model is tested focusing on restoration time of any machine. This paper identifies and predicts failures of heavy machines, thus facilitating the predictive maintenance scenario for effective working of the machine at all situations. This work is implemented by LSTM network model for gaining authentic results with numeric data which facilitates major cost savings and offers higher maintenance predictability rate.