Host company: Elder Research, Inc. (ERI) is a machine learning solutions provider with 26 years of experience delivering high performing predictive modelling solutions into complex client deployment environments.
The project: Fellows Faith Benson and Cyrille Feudjio worked alongside the team at Elder Research developing an unsupervised machine learning model to predict when parts are likely to fail within a hydroelectric power plant. The value of such a model is gained by applying its predictions to improve the maintenance and planning schedule of the power plant in order to improve its safety and overall efficiency in its operations.
During the fellowship Faith and Cyrille were presented with a large data set which was collected from sensors in the power plant at intervals. These sensors captured different data types such as temperature and pressure. They had a large job on their hands preparing the data which had 67 columns and almost 600 000 rows.
The outcome: Faith and Cyril were able to detect anomalies at certain times and also identify abnormal patterns which are likely to be the cause of the anomalies detected.
View their presentation on their work and solution below:
Jericho Mcleod, Data Scientist at Elder Research
“The value we are hoping to get is in their in-process work on a cutting-edge neural network implementation in Python. While the final product isn’t deliverable yet, they’ve made visible steps toward success, and in the modelling phase they’ve done this so faster than I expected. It looks as though we will finish this fellowship with an example of the model we can use to build it into pipelines”
We are pleased to announce that following the completion of the Fellowship, Elder Research have extended their contract with Faith and Cyrille so that they can continue working on the project.