IoT Analytics / Predictive Maintenance (Water Pump Shutdown Prediction)

Problem Statement

A small team is taking care of a water pump in an area distant from a town. There have been seven system failures in the past year. These failures caused severe inconveniences to many farmers and their families. The team couldn't see any pattern in the data when the system goes down, so they are not sure how to reach the root cause of the issue. They hope to understand from the data, which sensor or a group of sensors are causing the pump failure and closely monitor them.

Aim

Predict the next failure before it happens by monitoring the group of sensors that closely impact the working of the pump and replace the defective components before the next failure.

Feature selection using sensor readings (based on lowest weights)

Our Detailed Approach

The following phases of OSEMN Framework has been involved to develop the predictive model:

1. Obtain
2. Scrub
3. Explore
4. Model
5. Interpretation