Introduction of the position:
- Prove to Rijkswaterstaat that we can predict cable survey schedule intervals by applying analytics and machine learning to our existing cable survey data.
- Use data analytics and machine learning to identify cable sections that have remained stable in burial depth over time and are unlikely to require annual surveys. This would reduce costs, emissions, and internal workload, while improving planning and predictability.
- Analyse cable survey data to predict and locate areas for inspection, enabling a more data-driven, predictive and risk-based cable survey approach.
- Identify non-critical cable sections that consistently remain well above threshold levels and are unlikely to approach them in the near future, based on historical data and stable burial depth
How:
- Analyse historical data on Borssele cable and use ML algorithm to see if reburial that happened last year could have been predicted. Surveyed 3x a year, lots of data for analysis present.
- Identify non-critical cable sections that consistently remain well above threshold levels and are unlikely to approach them in the near future, based on historical data and stable burial depth.
- Centralize data through an intelligence hub.
Key Deliverables:
- Automating manual cable survey data analysis using analytics.
- Develop machine learning algorithm that is able to perform reburial need predictions
- Perform automated cable survey frequency prediction that deviates a maximum of 5% from the parallel manual analysis
- Predict reburial need for Borssele cable within 95% confidence interval based on historical data
- Start date on the 25th of November
- If you’re interested you can apply directly via e-mail and short motivation to Johanna.schacht@tennet.eu.