Skip to content

Use Case

Constructing and Learning from Seasonal Lock Usage Datasets

Description

Navigation decision support to the US Army Corps of Engineers (USACE) is provided by the Navigation Civil Works Decision Support Center (NDC). NDC provides information on locks through the Lock Performance Monitoring System (LPMS). Its primary purpose is to act as a planning tool for vessel operators.

The goal and context of this use case is to provide navigation decision support via more detailed descriptions about vessels operating in US waterways, and usage of locks by vessel type.

For example, the Public Lock Usage Report generated by LPMS is a summary of lock usage by vessel type, by year. Because these data are reported on an annual basis it is not possible to do any meaningful analysis of lock usage with respect to seasonality and weather including surface – water data (e.g., streamflow and gage height). The objective of this use case is to show how external variables can help decision support analysts to better understand variation in lock usage given datasets that are high-resolution with respect to both place and time. There is an opportunity to make this kind of analysis possible by systematically collecting data from the Corps Locks website

Target User(s)

  • USACE,
  • NDC,
  • DOT MARAD,
  • research centers and investigators (e.g., MarTREC),
  • vessel operators

Pre-TransMAP Hub Conditions

Lock Queue Report

“The Corps Locks website contains lock and vessel information derived from the United States Army Corps of Engineers Lock Performance Monitoring System (LPMS). The information contained here represents an every 15-minute updated snapshot of Freedom of Information Act (FOIA) data on U.S. flag vessels and foreign vessels operating in U.S. waterways that transited a Corps-owned or operated lock structure.”

Information is organized into nine (9) reports that include a Lock Queue Report with the following Selection Criteria: “Select a river system and a lock, then click the Submit button to view a report of the most recent vessel traffic reported at the lock.”

The report includes

  • Avg Processing in Minutes (Past 24 Hours),
  • Avg Delay in Minutes (Past 4 Hours),
  • Avg Delay in Minutes (Past 24 Hours),
  • Last Updated (EST),
  • and Most Recent Lockages.

They are described with the following columns:

  • Vessel Name,
  • Vessel No,
  • Direction,
  • Number of Barges Proc,
  • Sol Date,
  • End of Lockage,
  • Timezone,
  • Delay Minutes,
  • Seq No.

“A flashback query allows you to view data as it existed at a previous point in time. As of [0, 1, 2, …, 99] minutes ago.”

Public Lock Usage Report

The currently available, downloadable report has File Name: Lock Usage CY 1993-2019.XLXS

Its first sheet is Table of Contents – Locks by Waterway. The second sheet contains tables for All Waterways, each individual waterway, and every Waterway Lock. Data are arranged inside of tables such that columns represent calendar years, and rows represent the following usage variables

  • Average Delay
  • Average Processing Time
  • Barges Empty and Loaded, Commercial Vessels, Flotillas, and Lockages
  • Non-Vessel Lockages
  • Non-Commercial Vessels, Flotillas, and Lockages
  • Percent Vessels Delayed
  • Recreational Vessels and Lockages
  • Total Vessels and Lockages

Post-TransMAP Hub Conditions

By systematically collecting data from the Lock Queue Report it would seem possible to create a new version of the Public Lock Usage Report such that its basis in time is high-resolution compared to the annual observations currently available. Even quarterly data would allow the decision support analyst to begin considering seasonality and controlling for weather in models of lock usage

Variables that appear in similar forms across the Lock Queue Report and Public Lock Usage Report are the following:

  • Average Processing
  • Average Delay
  • Lockages
  • Barges
  • Delay

It may also be possible to add information about Direction to future, more detailed datasets about lock usage. It is an objective of this use case to illustrate how any lock usage data collected on a better-than-annual basis could be linked with weather data. Even without weather data the ability to adjust Lock Performance to continuous rather than seasonal operation would represent a marked change in navigation decision support provided by NDC.

Advantages & New Capabilities

Users will have access integrated, spatiotemporal datasets relevant to lock usage and weather. In theory the decision support analyst or user could choose among levels of resolution with respect to both place and time. Examples of place are lock, division, district, and waterway; examples of time are quarterly, monthly, weekly, and daily. Users would receive resulting sheets of data for analysis – and visualizations – related to lock usage, seasonality and weather. Based on this use case, TransMAP may also recommend strategies for dealing with the problems of missing data and interdependence among variables, described below.

Challenges

There are at least two significant challenges associated with analyzing lock usage data:

  1. Lock usage data tend to be incomplete. In other words, we have the problem of missing data in the Public Lock Usage Report. It is possible this problem could be made less severe by systematically collecting data from the Lock Queue Report, but maybe not.
  2. Lock usage data (and weather data for that matter) tend to be interdependent. In other words, we may have the challenge of predictor variables influencing each other, which can make it difficult to learn from data.

These problems are not easily overcome, but they can be dealt with by a knowledgeable decision support analyst. We will attempt to address these challenges by example, in the completed use case.

Frequency of Use

Frequency of use will likely depend on the time basis for creation of new datasets. In a way, the decision support analyst currently has a full year to update relevant planning tools, as the Public Lock Use Report is updated annually. A quarterly report based on Lock Queue data would suggest that frequency of use should be at least four times per year.

Contacts

Justin Chimka

Associate Professor, Department of Industrial Engineering, University of Arkansas, Fayetteville

Center for Advanced Spatial Technologies (CAST)


Last update: 2021-06-30