CAIT 10: Inland Marine Transportation Analytics, Phase II
This research proposal represents a continuation (second phase) of research previously initiated at Vanderbilt University. One of the tasks in particular, the use of the OMNI system, was not able to be completed as the Corps of Engineers transitioned to a new lock performance system. Under previous contracts, Vanderbilt researchers have studied inland marine transportation casualties and the use of AIS data and other real-time data sources as a source of better trip data. The marine casualty research attempted to identify the most hazardous locations of the inland waterway network. Casualties, for purposes of this research, are considered groundings (vessel hits bottom), allisions (vessel hits a fixed object, such as a lock or bridge) and collisions (between 2 or more moving vessels). The approach and visualization techniques were successful for understanding casualty frequencies at selected locations. However, the absence of high-resolution trip data precluded the calculation of casualty rates (e.g., collisions per million trips, etc.). Correlation of casualties with weather and river stage information was also identified as a critical research need. These needs led to the research conducted in 2011, where AIS was studied as a potential source of vessel trip data. During 2011, Vanderbilt researchers archived and analyzed 6 months’ of AIS data from the Paducah, KY region. The findings of this research included that AIS has great potential for serving as a source of trip data. However, as AIS is not required above Baton Rouge, only 1 in 3 towboats transiting the Paducah case study area were found to have transmitting AIS equipment installed. The research also discovered additional data sources published by the U.S. Army Corps of Engineers (USACE) that would be useful in benchmarking vessel traffic on the inland waterways. The proposed research is intended to advance the previous work done by Vanderbilt by applying analytics and visualization techniques to the inland marine transportation areas studied. This research should also produce better accident rates for the Lower Mississippi River between New Orleans and Baton Rouge through the use of more reliable traffic data captured by the AIS receiver.
The objectives of this research are expressed in the following three focal areas:
- Study the relationship between historical inland marine casualties and severe weather and high water events – this research will tag each marine casualty in the database (dating from 1980-present) with the weather and river conditions during the time of the casualty. This information is expected to include air temperature, precipitation, cloud cover, visibility and river stage.
- Explore the integration of real-time data sources for inland marine transportation operational analytics – this task will include the investigation of various data sources (primarily OMNI reports and weather/river condition information) for real-time integration. The feasibility of integrating this information in a web-based, geographical environment will be explored. Potential metrics include waterway congestion, lock queues, and average tow size during high water conditions, among others.
- Analyze AIS data captured at Reserve, Louisiana facility – Reserve is a city about halfway between New Orleans and Baton Rouge. This location has been selected since AIS is required to be carried by all vessels in the area (below Baton Rouge on the Lower Mississippi River) and there is a more diverse traffic mixture (deep sea and inland vessels) than was observed at the Paducah location. The data capture and analysis routines developed during the Paducah AIS project will be improved so that results are available with a much quicker turnaround. It is also anticipated that since AIS carriage is mandatory in this area, the information captured may be more comprehensive (cargo information may be populated more frequently) and of higher quality.
Key Research Activities:
The following tasks will be performed as part of this research:
- Update marine casualty data from U.S. Coast Guard (USCG) MISLE system – the most recent casualty data from the USCG will be obtained. There may be additional attributes, as the current database (produced during previous CAIT research) covered three different USCG reporting systems from 1980-2010: CASMAIN, MINMOD, and MISLE.
- Relate historical weather and river stage information to casualties – as mentioned previously, each casualty will be tagged with weather and river stage conditions (based on the casualty location and date/time). The relationship between severe weather, tow size (if the data is available), geography and casualties will be explored.
- Identify and select sources of real-time inland waterway data (from OMNI reports and USGS river gage information systems) – this research task involves studying the OMNI report XML services and identifying the sources of real-time weather data (including map services, warnings and river flow/stage conditions) for inclusion in the analytics portion of the proposed research.
- Modify AIS data routines to capture and analyze data from Reserve, LA – during the data analysis phase of the Paducah AIS project, several sampling routines (e.g., one position every 15 minutes, etc.) were explored. All of the routines will be combined to run in tandem once log files arrive at the server. This way, the published results are never more than four hours old.
There are many benefits to this research. Benefits of the three focal areas are presented below:
- Improved understanding of marine casualties and their root causes will lead to safety benefits by determining if certain severe weather conditions lead to more casualties. This will be studied by region, river, and direction, by virtue of using geographic information systems (GIS) technology to conduct the analysis. For example, it may be advisable to reduce a fleet’s tow size in the face of approaching severe weather.
- Applying analytics and business intelligence techniques to a combination of real-time datasets is expected to lead to higher quality information on inland marine transportation conditions and events. With improved information on lock performance and waterway congestion, for example, fuel consumption savings may be realized by slowing tows which are approaching a congested lock.
- The potential of using AIS data to quantify vessel trips was established in the Paducah AIS study. Difficulties were caused by the location of the antenna at the junction of three river systems and significant blind spots (due to elevation). The Reserve, LA location has a simpler river configuration and less variations in elevation, so the data is expected to be easier to analyze. Additionally, by studying the data received in an area of required AIS carriage (Reserve, LA), researchers expect to see differences in the quality of data and improve routines to process the data for more timely results with ad-hoc querying capabilities.
Dr. James Dobbins is the principal investigator of this project.