Technology Stack
Predictive Analysis
Challenge
A transport company wanted a way to detect wear patterns in overhead wire on electric train lines with a focus on identifying specific wire wear over time and predicting wire wear and useful life moving forward.Â
Solution
Blackbook’s solution focused on pivoting the focusing to on contact between the pantograph and the wire. Using the Staggering Peak Model, Blackbook was able to identify mid points, which helped reduce the amount of exceptions. Focusing on finding mid-points removes noise from faulty sensors and is able to be a more robust predictor. As a result the Staggering Peak Model detects change in the staggering over time and can detect threshold exceptions such as wire wear.
Outcome
From the model, the client can now better predict wire wear of the overhead wires on the tracks, enabling them to be more aware of faults and act quicker on repairs.Â