Skip to content

This project contains an implementation of a decision tree model for maintenance and deterioration

License

Notifications You must be signed in to change notification settings

kaleoyster/decision-tree-nbi

Repository files navigation

🌉 Predicting Bridge Maintenance Using Decision Trees 🌟

The repository is implementation of prediction models to identify future bridge deterioration and maintenance using the existing National Bridge Inventory (NBI) dataset. This project is a part of the Bridging Big Data Group and SMARTI at University of Nebraska, Omaha.

[Interactive visualization of Results by Ashley Ramsay](https://repairs.ricks.io/)

🎯 Background

  • Bridges are critical infrastructure whose maintenance is a key responsibility for both government and private organizations.
  • Costly to build and repair, bridges are valuable assets that can benefit from improved predictive maintenance.
  • In this work, we present a novel, straightforward, non-parametric method for doing predicting future maintenance and identifying influential factors to predict future maintenances of the bridges.

💪 Challenge

  • In the NBI dataset, it is a challenge to identify maintenance patterns. Because, the reconstruction_year does not provide an valuable information about the type of repair or reconstruction done.
  • Moreover, for every bridge time-line, the improvement in the condition ratings often does not map with the data in reconstruction_year.

🎯 Objective

  • The objective of this research study is to develop a methodology for computing bridge maintenance frequency and based on existing timelines of the available data predict future time-line by accounting for improvement in the condition ratings. Methodology

🧪 Solution

  • To address this challenge we can use bridge intervention matrix, that utilizes the bridge intervention matrix for deck to identify various types of intervention depending on the probability of the transition.

Deck Bridge Intervention Matrix

  • The project contains several models which are broadly specified under two categories:
    • Deterioration models: In general, deterioration models predict the future deterioration of bridge conditions.
    • Maintenance models: In general, maintenance model predict the future maintenance of bridge conditions. There are three main components for which maintenance models can make predicitons:
      1. Substructure
      2. Superstructure
      3. Deck

👉 References

Document Documentation type Description
Quickstart Concept An overview and guide to setup this project
Methodology Concept, Task Simplest possible method of implementing your API
Functions Reference List of references for the functions used
Related Projects Reference List of projects related to this repository

Contact

About

This project contains an implementation of a decision tree model for maintenance and deterioration

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published