AI Usage in Infrastructure Maintenance
Advisor: Ms. Olender
Internship Location: NYC Department of Transportation
Onsite Mentor: Zhihua Yi, Administrative Engineer
Bridges are essential to transportation systems, yet thousands in the United States are aging and require continuous monitoring to remain safe. As inspection demands grow, artificial intelligence is emerging as a potential tool for identifying structural risks earlier and improving maintenance planning. This project will investigate how artificial intelligence and predictive data analysis could improve maintenance strategies for bridge infrastructure. The study will focus on whether machine learning and large infrastructure datasets can help identify deterioration patterns and potential structural issues before they become critical. Using data from infrastructure databases such as the National Bridge Inventory, I will examine trends in bridge condition ratings, deterioration patterns, and maintenance history. Additional insight will be gained through an internship with an administrative engineer at the New York City Department of Transportation, where I will observe infrastructure inspection practices and learn how bridge data is collected and used in real-world maintenance decisions. Through this research, I expect to develop a framework for analyzing bridge condition data and evaluating how AI-based models could detect patterns associated with structural deterioration. The analysis is expected to demonstrate how predictive systems may help engineers identify maintenance needs earlier and prioritize repairs more effectively. If successful, the research could demonstrate how combining traditional inspection methods with predictive data analysis may improve bridge safety and help transportation agencies allocate maintenance resources more efficiently.
Project Blog
Week 0: Introduction
Bridges are essential to the transportation systems we rely on every day, yet thousands across the United States are aging and require continuous monitoring to remain safe. There are over 623,000 bridges throughout the U.S. and 6.8% of those are currently rated as “poor”. That means over 40,000 bridges are not safe enough for too much traffic, but there are 178 million trips over those bridges every day. Much of the problem is attributable to their age: over 40% of the nation’s bridges have gone over 50 years without proper renovations (ASCE Infrastructure Report Card). Most were not constructed to last that long; our expectation that a bridge will provide service for 50 to 70 years is based on their typical longevity, and yet the age of America’s bridges continues to increase.
As we continue to struggle to keep up with increasing inspection and maintenance demands, funding for bridges poses a serious problem as well. Currently, the rehabilitation costs of bridge systems are estimated at $191 billion, but even with additional federal funds, the number of bridges rated in “fair” condition exceeds the number rated in “good” condition (ASCE Infrastructure Report Card). Previously unnoticed structural deterioration could escalate from requiring routine maintenance to becoming a serious public safety concern (such as the collapse of the Fern Hollow Bridge in Pittsburgh). The National Transportation Safety Board determined that serious gaps in maintenance and oversight ultimately resulted in the bridge collapsing, even though the bridge remained open for a substantial amount of time after inspectors determined it was “safe for travel.”
As AI continues to evolve, it may play an important role in addressing this challenge sooner in the process. A number of studies have demonstrated this and have reported promising results. One study found a machine learning model to predict with 94.23% accuracy from bridge inventory data and reduced the number of bridges stated to require scheduling for inspection by nearly 30% (MDPI). This level of increased efficiency has the potential to result in transformative benefits on a larger scale.
My research will explore the application of machine learning to large scale infrastructure data and whether it will identify patterns of deterioration and/or potential structural issues before they reach critical status better than out current models. I will analyze existing data to review trends in bridge condition ratings, identify common patterns of deterioration, and look at their maintenance history so predictive models can identify structures that may need maintenance sooner than would typical engineering practices.
In addition to analyzing the data, I will also gain experience through a hands-on internship with an administrative engineer at the NYC Department of Transportation. In this role, I will witness firsthand how infrastructure inspections are conducted, and I will become familiar with how bridge data is collected and used to inform real-world maintenance decisions. This information will be critical to ensure my research is applicable to how these systems actually work in practice.
Department of Transportation: https://www.nyc.gov/html/dot/html/home/home.shtml
Week 1: Getting Started
Hi everyone! This week, I made a lot of progress on research.
First, I wanted to see how bridge inspections/maintenance work currently in the United States. I looked at the Federal Highway Administration Bridge Preservation Guide and the American Society of Civil Engineers Infrastructure Report Card, and the statistics were surprising.
According to the ASCE Report Card, U.S. bridges were given a C grade. 42% of U.S. bridges are at least 50 years old, and an estimated $125 billion national backlog for rehabilitation remains. After looking at the rate of investment for bridge repairs, I realized that meant that it will take approximately 50 years just to fix all of the repairs that need to be made today. That is not a small problem: it’s a systematic one.
The Bridge Preservation Guide by the FHWA was perhaps the most beneficial thing I read this week. This guide will take you through the lifecycle of a bridge from a maintenance perspective: determining at what point an agency determines that they need to intervene, what kinds of treatments are available, and how to analyze the cost of taking action early versus waiting too long. I learned that bridge maintenance can be classified into two basic types: preventive maintenance and reactive maintenance. Preventive maintenance is proactive, focusing on applying treatments to bridges in good condition before they have serious deterioration. It is similar to changing your oil on schedule prior to an engine failure rather than after. Reactive maintenance only occurs after damage has occurred to the structure, making it a much more expensive and risky option for the agency. What I took away from the guide is that the system knows what good maintenance practices are supposed to be, but due to an overwhelming number of bridges and limited financial resources available to agencies, it is almost impossible for agencies to consistently stay ahead of the curve in bridge maintenance.
I'm looking forward to seeing what the next couple of weeks will bring. I'm hoping to go on-site and witness the actual bridge maintenance and inspections of the bridge structures first-hand. With respect to research, I will be looking into how to incorporate artificial intelligence and machine learning into existing bridge inspection systems, as well as predictive modeling that already exists in this space, and how those models compare to real bridge condition data. Ultimately, I hope that this work results in my own established framework that can connect the traditional methods of maintaining bridges with the possibilities offered by the modern data analytics landscape.
Below, I attached my Senior Project Proposal for you guys to look at if you are interested.
Updated Philip_Senior Project Full Proposal
Work Cited:
American Society of Civil Engineers. 2021 Infrastructure Report Card: Bridges. ASCE, 2021, www.infrastructurereportcard.org/cat-item/bridges-infrastructure/.
Federal Highway Administration. Bridge Preservation Guide: Maintaining a State of Good Repair Using Cost Effective Investment Strategies. U.S. Department of Transportation, FHWA, www.fhwa.dot.gov/bridge/preservation/guide/guide.pdf.
Week 2
Hey everyone! I spent this past week transitioning my focus from the condition of the bridge infrastructure in the US to potential technologies that could help fix this.
After discovering last week about the extensive flaws currently embedded within existing methods for maintaining bridges, some of which will potentially take decades to overcome and thus jeopardize public safety, I believed that it would be a logical next step for me to examine what AI can contribute towards improving access to the data and conducting analyses on the maintenance of bridges. This week was primarily devoted to learning about foundational concepts (machine learning, data modelling, sensor networks, and AI-enabled structural health monitoring systems). I wanted to make sure I had a solid grasp on all of these before diving into the research literature.
After laying out the basic concepts, I moved into the various academic journals and case studies examining how artificial intelligence is currently being applied to infrastructure testing. Through machine learning algorithms, AI can create very sophisticated computer algorithms for analyzing and predicting potential problems based on large sets of real-time data, coming from a variety of sensors that have been placed within bridge structures. These are meant to continuously monitor changes in conditions (e.g., stress levels, vibrations, etc.) and will be used to develop predictive systems for predicting when these changes will require some type of maintenance work to be done before anything has become seriously damaged.
One study I came across applied a machine learning model directly to National Bridge Inventory data and reached a prediction accuracy of 94.23%, while cutting the number of bridges scheduled for inspection by nearly 30%! In another study done in Taiwan, researchers analyzed over 27,000 records of past inspections and concluded that a Random Forest type of model (a machine learning algorithm that makes predictions by building a large number of decision trees and combining their results) would return a prediction accuracy of 88% in predicting which individual components of a bridge were likely to be deteriorating. This outperforms many more traditional types of algorithms in the process, and results in the ability to apply a data-driven approach to prioritizing maintenance on a bridge, thereby allowing for much greater efficiencies to be identified in the inspection and maintenance of bridges.
From everything I've seen throughout this research, AI will not take the place of Engineers (I hope), but rather give them more accurate and detailed information. The best way to do this is through a "human in the loop" framework: that's when outputs from a model are combined with field observations and engineering judgement. That distinction is critical, and it's going to form the framework for this project. There is no question about whether the technology will be there. It's whether we can utilize this technology responsibly and effectively as we develop actual infrastructure systems.
Work Cited
Contreras-Nieto, Cristian, et al. "Bridge Maintenance Prioritization Using Ordinal Optimization and Machine Learning." Automation in Construction, vol. 104, 2019, pp. 243–256.
Jiang, Yi, et al. "Predicting Bridge Deterioration Using Machine Learning: A Case Study on the National Bridge Inventory." Journal of Infrastructure Systems, ASCE.
Sagi, Omer, and Lior Rokach. "Explainable Decision Forest: Transforming a Random Forest into an Interpretable Tree." Information Fusion, vol. 61, 2020, pp. 124–138.
Week 3
Hi everyone! This week, I continued building on my AI research, this time zooming out from bridges specifically to look at how artificial intelligence is being applied across transportation systems more broadly.
I looked at various journals and case studies on the application of artificial intelligence technology to railways, highways, and transit systems. A good resource for me was a survey article published in Sensors called "A Survey of Artificial Intelligence Predictive Maintenance Methods for Railway Infrastructure." It reviewed machine learning and deep learning techniques (including neural networks, support vector machines, and random forests) for analyzing things like the geometry of the track, vibration levels, and images of the infrastructure. What stood out to me most after reading these articles was the similarities among the many predictive maintenance methods as they relate to the infrastructure on railways; the fundamental principles for measuring conditions of railway infrastructure are very similar to those for measuring bridge conditions.
I found an article in Scientific Reports about an explainable machine learning framework developed for the Metro of Porto, Portugal, which created a real-time predictive maintenance pipeline that processes sensor data, flags anomalies, and provides natural language explanations of these outputs to make machine decision-making understandable to engineers and not just data scientists. I was fascinated by the concept of explainability. A model that can tell you why it flagged a structure as at-risk is far more useful to a real engineer than one that just spits out a number.
What all of these articles reinforced is something I keep coming back to: the lack of visibility into asset condition, sparse monitoring, and manual analysis hinders traditional maintenance practices. Artificial intelligence will eliminate all three of these obstacles and allow for continuous, data-driven monitoring of all types of maintenance assets, including but not limited to track and bridge assets.
Outside of research, I have begun to look into potentially volunteering at transportation museums or getting direct access to bridges to better understand the physical side of infrastructure maintenance. There are many excellent sources of transportation/infrastructure exhibits in transportation museums throughout New York City, such as the New York Transit Museum, the New York City Department of Transportation's historic archives, and the Intrepid Sea, Air & Space Museum. These museums would provide a tremendous resource to develop my familiarity with the physical presence of infrastructure, as well as how to work with and understand the actual environment (scale, how each component relates to one another) that inspectors encounter while working on the infrastructure. Access to these structures would only improve my ability to carry out my analysis and make my final framework a true representation of reality.
Work Cited
Gihub.org. "Sensors and Machine Learning for Predictive Maintenance." Global Infrastructure Hub, 2020, www.gihub.org/infrastructure-technology-use-cases/case-studies/sensors-and-machine-learning-for-predictive-maintenance/.
Meira, Jorge, et al. "Data-Driven Predictive Maintenance Framework for Railway Systems." Integrated Computer-Aided Engineering, vol. 30, no. 3, 2023, journals.sagepub.com/doi/abs/10.3233/IDA-226811
Pérez-Pérez, Enrique Jose, et al. "A Survey of AI-Enabled Predictive Maintenance for Railway Infrastructure: Models, Data Sources, and Research Challenges." Sensors, vol. 26, no. 3, 2026, www.mdpi.com/1424-8220/26/3/906.
Veloso, Bruno, et al. "An Explainable Machine Learning Framework for Railway Predictive Maintenance Using Data Streams from the Metro Operator of Portugal." Scientific Reports, 2025, www.nature.com/articles/s41598-025-08084-1.
Week 4
Hi everyone! This week, I had my very first meeting with my advisor, Zhihua Yi, which helped clarify my vision for this project. I have spent the past couple of weeks building my understanding of AI and predictive maintenance, and researching how these solutions have been tested or are being tested across transportation systems. As a result, I've decided to shift my focus toward something more specific and more directly impactful: how AI can get the most accurate and detailed structural information possible into the hands of the contractors who carry out major bridge repairs.
The current shift in focus for me came from a conversation I had with Zhihua during our first meeting. I learned that although AI is an exciting tool to potentially monitor all aspects of the world, what really matters at the moment is how it is actually used for the better. For example, when a contractor arrives on-site to repair a bridge, they won’t be able to complete the repair based solely on the information they received before arriving at the site. Right now, there are numerous problems when an inspector provides a report on the state of the bridge, such as it's missing critical data or is based purely on their visual inspection without taking into account any deteriorated structural components below the surface of the roadway.
Research supported this claim; traditional inspection methods (e.g., visual inspections and manual load testing) are inherently susceptible to significant flaws and inaccuracies due to human error, are labor-intensive, typically only provide surface evaluations, and often fail to find underlying structural problems. According to an article in the International Journal of Bridge Engineering, Management & Research, a New York State study determined that some bridge components (e.g., parapet walls) were rated differently by different bridge inspectors, meaning that two engineers could obtain totally different conclusions. For contractors who need reliable estimates of the amount of repair work required to get their job finished safely/substantially, inconsistent evaluations can be dangerous, potentially depriving them of the ability to evaluate an adequate data set on which to base their repairs.
There's a major difference between AI and traditional maintenance in terms of costs and efficiency. Traditional maintenance reacts to issues as they arise at or after the point at which they are noticeable and/or require expensive repairs, which is expensive and risky because it is possible that critical issues might go unnoticed, or that an issue goes unsupported in a facility until it is too late. AI solutions radically change this by providing early detection of critical structural issues, which helps organizations avoid costly emergency repairs, avoid periods of downtime, and reduce or eliminate catastrophic failures. AI solutions can also optimize maintenance schedules, thereby reducing the overall number of routine inspections and reducing costs over time. One study found that a facility that implemented AI-based safety protocols was able to reduce its workplace incidents by 75% and also experienced an increased compliance rate through automated inspections and monitoring of its facilities.
It’s a dangerous time in the world of contracting when contractors are so restricted by the lack of accurate information when attempting to assess the structural condition of a bridge. In the future, my primary emphasis is to utilize AI to eliminate that gap, providing the most specific, accurate, and precise structural data available to the individuals charged with performing major repairs on bridges, so that every single contract will be planned and performed as well as possible with the greatest information available about what they’re getting into. That is how AI will make a direct impact to enhance the safety of bridges.
Works Cited
Encardio. "AI-Driven Bridge Monitoring: Enhancing Safety & Efficiency." Encardio Rite, 2025, www.encardio.com/blog/ai-driven-bridge-monitoring-enhancing-safety-efficiency.
Harle, S.M., et al. "Advancing Bridge Infrastructure Management through Artificial Intelligence: A Comprehensive Review." International Journal of Bridge Engineering, Management and Research, 2025, ijbemr.org/index.php/ber/article/view/45.
Netguru. "How AI Predictive Maintenance Cuts Infrastructure Failures by 73%." Netguru, 2025, www.netguru.com/blog/ai-predictive-maintenance.
Wang, Naiyu, et al. "An Improved Inspection Process and Machine-Learning-Assisted Bridge Condition Prediction Model." Buildings, vol. 13, no. 10, 2023, www.mdpi.com/2075-5309/13/10/2459.
Week 5
Hi everyone! This week, I narrowed my research focus significantly, and it felt like a big step in making this project feel real.
As I have spent the past several weeks learning about general concepts related to AI, predictive maintenance, and bridge inspection practices across the United States, I have decided it is time to focus my efforts on a case study of a specific bridge. I chose the Hamilton Avenue Bridge (BIN: 2240231), which is a movable bascule bridge (drawbridge) that spans the Gowanus Canal, in Brooklyn. I pass by this bridge all the time, as it’s very close to Basis and near where the train stop is (not to be confused with the 9th Street Bridge). I am familiar with the various traffic patterns it experiences and have seen the level of usage it receives every day. Researching a bridge that I know seemed like the best way to make my project as relevant and specific as possible.
There is much to be learned from the history of the Hamilton Avenue Bridge, and as such, there are many reasons to study it. The current drawbridge portion was built in August 1942 and is completely below the elevated Gowanus Expressway.
The Hamilton Avenue Bridge underwent reconstruction in 2008 & 2009 by the DOT. This included new electrical and mechanical operating systems, a new fendering system, and navigation lights for the marine traffic. Even with all of these improvements, the original construction of this bridge is now over 80 years old, and it carries a massive amount of weight.
The Hamilton Avenue Bridge has eight lanes of traffic that cross the Gowanus Canal beneath the elevated Gowanus Expressway, making it one of the largest movable-bridge crossings in Brooklyn. According to the NYC DOT’s Annual Condition Report for 2024, B.I.N. 2240231 is currently rated 6.0 (NBI) based on its last inspection (August 2023) and is rated as Fair. A Fair rating on a movable bridge with this heavy volume of traffic creates a scenario where predictive data analytics could have the most significant impact.
What makes this bridge especially compelling as a case study is its environment. The Gowanus Canal is currently being cleaned up by the U.S. Environmental Protection Agency (EPA). During the clean-up process, the canal will require opening the Hamilton Avenue Bridge to allow clean-up boats to cross under the bridge. There will also be times when the bridge will be closed to vehicular traffic so that routine maintenance can take place, resulting in delays and detours. Therefore, the Hamilton Avenue Bridge must deal with two additional problems in addition to age and traffic load: the corrosive conditions of the waterway and the structural wear caused by chemicals in the waterway, which will likely accelerate more rapidly than anticipated based on current inspection cycles. The Gowanus Bridges have been rebuilt over twenty times and have malfunctioned so frequently that the reason for the construction of the elevated expressway above was due to the bridge's multiple malfunctions. The history of mechanical failures associated with the Gowanus Bridges makes the case for developing improved predictive tools for bridge inspection and maintenance more compelling.
As I move ahead with developing a framework using the inspection data, condition ratings, and maintenance history for the Hamilton Avenue Bridge, I will analyze how Artificial Intelligence can assist in providing more accurate and detailed structural information to contractors who will perform significant repair work to the bridge, prior to visiting the bridge in person.
Work Cited:
Bklyner. "A Brief History of the Bridges Over the Gowanus Canal." Bklyner, 2019, bklyner.com/a-brief-history-of-the-bridges-over-the-gowanus-canal/.
Gowanus Superfund. "Gowanus Canal Remediation Overview." Gowanus Superfund, gowanussuperfund.com/.
New York City Department of Transportation. Bridges and Tunnels Annual Condition Report, 2024 Edition. NYCDOT Division of Bridges, 2024.
NYC DOT. "The Bridges of the Gowanus Canal." NYC.gov, nyc.gov/html/dot/html/infrastructure/bridges-gowanus.shtml.
South Brooklyn Network. "Gowanus Canal." South Brooklyn Network, southbrooklyn.com/2020/01/01/gowanus-canal/.
Star-Revue. "The Bridges of Gowanus." Red Hook Star-Revue, 2020, star-revue.com/the-bridges-of-gowanus/.
Week 6
This week, I decided to switch up from last week’s bridge and focus on one I am even more familiar with.
I've taken a closer look at bridges over the Gowanus Canal and decided to move on from the Hamilton Avenue Bridge, focusing my case studies on the Third Street Bridge (B.I.N 2240250). If you have ever been to the Whole Foods on 3rd St. in Gowanus, you most likely have crossed the Third Street Bridge and found it right next to the entrance to the parking lot. I walk past this bridge all the time, and the deeper I researched it, the more interesting it is from a maintenance perspective, as it has a low inspection rating (<5) and is one of the more complicated bridges to maintain in Brooklyn.
A few things make it stand out. It originally opened in 1905 but underwent major repairs in 1986. This work included replacing the mechanical and electrical systems and performing a partial structural rehabilitation of the bridge: the majority of the structure itself is now more than 120 years old! It currently has a National Bridge Inventory Weight Rating of 4.65 (rated as Fair), meaning it is just above the conditions to be rated as “poor” (less than 4). I also selected this bridge due to its simple Scherzer rolling-lift mechanism as opposed to more elaborate drawbridge designs; therefore, I believe that a predictive model built around this bridge will be more applicable to the average older movable bridge than it will be to a unique and complicated structure.
The location adds another layer of difficulty. In June 2024, the EPA will begin the Superfund cleanup of the middle section of the canal at an estimated cost of $369 million and will take multiple years to complete. The Third St. Bridge is located precisely at the boundary of this cleanup. Frequent bridge closures will be required while the structures along the canal are inspected and repaired to obtain full structural support during the continuing dredging process. Therefore, the bridge is subject to chemical exposure, continual vessel traffic, and mechanical stresses on an already aged structure.
In the future, I will attempt to develop a predictive simulation that will utilize the bridge's inspection history and condition ratings, as well as maintenance data, to determine how an AI system would recognize deterioration trends and alert to any structural hazards before they become larger problems. A more complex and industry-tuned version of this would ideally develop a means for contracting agencies to receive this information prior to starting any major repairs on the bridge. This is the clearest direction I have taken thus far with respect to this project.
Work Cited:
Bklyner. "A Brief History of the Bridges Over the Gowanus Canal." Bklyner, 2019, bklyner.com/a-brief-history-of-the-bridges-over-the-gowanus-canal/.
Gowanus Canal Conservancy. "Superfund." Gowanus Canal Conservancy, 2024, gowanuscanalconservancy.org/superfund/.
Gowanus Superfund. "FAQ." Gowanus Superfund, gowanussuperfund.com/faq/.
Michael Minn. "The Gowanus Canal: Third Street." michaelminn.net, michaelminn.net/newyork/mobility/gowanus-canal/third-street/index.html.
New York City Department of Transportation. Bridges and Tunnels Annual Condition Report, 2024 Edition. NYCDOT Division of Bridges, 2024.
U.S. Environmental Protection Agency. "Gowanus Canal Superfund Site Profile." EPA, cumulis.epa.gov/supercpad/SiteProfiles/index.cfm?fuseaction=second.cleanup&id=0206222.
Week 7
The most significant accomplishment of this week would have to be developing a full outline for the research paper after meeting with my placement mentor. The outline of the research has made me realize that I would like to expand my focus just beyond structural monitoring. There is tremendous potential for AI in bridge maintenance, beyond just crack and corrosion detection; areas such as data organization and management of maintenance records, inspection scheduling, and formatting and disseminating information to the appropriate parties. I will also explore all of these areas in the research paper.
In addition to the initial project, I am going to do a second case study of the Pulaski Bridge's Jackson Avenue approach span (B.I.N. 2240639) in Queens, which connects 11th Street (in Queens) to McGuinness Boulevard (in Brooklyn) over Newtown Creek. The bridge has six travel lanes, as well as a pedestrian walkway; it terminates at a traffic light at Jackson Avenue (in Queens), subsequently continuing as 11th Street. This section of the bridge (Jackson Avenue) consists of fixed, non-movable approach span structures, meaning that there are no drawbridge mechanisms (e.g., counterweights or hoists) involved in its design/operation. Essentially, this structure is about as simple as a bridge can be built structurally. The NYC Department of Transportation (DOT) has rated this section of the Pulaski Bridge a fair rating of 5.2 NBI-W according to the 2024 Annual Condition Report.
For this reason, I am establishing it in my report. The Third Street Bridge is a classic example of a bridge that has a long history and complex mechanical elements (the bridge is over 120 years old and is located in a Superfund environment). This offers a more complex and potentially more environmentally impactful situation to study. In contrast, the Jackson Avenue Bridge is a high-volume, fixed-span, standard bridge that has few mechanical variables, making it a much simpler example from which to demonstrate how AI-based deterioration modeling can be performed easily and consistently. This combination of case studies allows the resulting research to have significant variability and produces a final model more generally useful than the individual case studies on their own.
Work Cited:
East Coast Roads. "Pulaski Bridge." East Coast Roads, www.eastcoastroads.com/states/ny/nyc/pulaski.
Michael Minn. "The Gowanus Canal: Third Street." michaelminn.net, michaelminn.net/newyork/mobility/gowanus-canal/third-street/index.html.
New York City Department of Transportation. Bridges and Tunnels Annual Condition Report, 2024 Edition. NYCDOT Division of Bridges, 2024.
NYC DOT. "The Bridges of Newtown Creek." NYC.gov, www.nyc.gov/html/dot/html/infrastructure/bridges-newtown.shtml.
Wikipedia. "Pulaski Bridge." Wikipedia, en.wikipedia.org/wiki/Pulaski_Bridge.
Week 8
Hi Everyone! This week I started writing my research paper!
My outline is now solid thanks to a research paper titled “Data Driven Bridge Deck Deterioration Modeling and Maintenance Intervention Scheduling,” which I am using as a guidance source for this project. This paper creates a new way to improve how bridges deteriorate are modeled through accounting for the problem of too much censoring found within the National Bridge Inventory (NBI) dataset of inspection data. Additionally, this research paper compares and contrasts AI-based methods with existing models used by State Departments of Transportation.
By reading this paper, I will learn how other research studies in an academic arena are structured, including how others define their issues (or state their problems), present their data, develop methodologies to support their findings, and place their findings in context. This will enable me to easily create my outline or template for writing my own paper.
It is relevant to my own work, as it discusses the limitations of traditional bridge deterioration models that rely on subjective judgment from inspectors or rely upon an empirical model, resulting in data encoding errors, and these kinds of problems will then cause reactive maintenance to occur only after significant deterioration has occurred, making this the exact reason for which I have based my research on. The paper supports my research foundation by giving me access to peer-reviewed research that provides a model I can use as I develop my own methodology and research findings.
Work Cited:
Kumar, Deepak. Data Driven Bridge Deck Deterioration Modeling and Maintenance Intervention Scheduling. CUNY Academic Works, 2024
Week 9
This week I began writing the Background, literature review section of my research paper. This section of the paper is important because it sets the foundation for the rest of the paper (the scope of bridge infrastructure in the US; the limitations of standard inspection methods; how AI will change the way we inspect our bridges). It also requires a great deal of attention to detail in order to ensure that enough context is provided without overwhelming the reader with too much information before the real substance of the paper begins.
Besides writing, I also began thinking about the presentation, specifically converting the dense nature of a research paper into an interesting and non-technical format that will engage an audience of non-technical people. Emphasizing certain points, omitting the unimportant parts and developing an appropriate manner of illustrating the research presented will be a whole different challenge.
Week 10
This week has been very productive for me as far as my writing goes. The literature review and problem statements are mostly written, and the general framework of the paper feels good and connected. The research I've conducted in the last several weeks has finally evolved into something that seems substantial and like an actual piece of writing, so I feel good about how it all comes together.
Now that the paper is on the right track, it's time for the presentation. This is going to be its own challenge, translating from a dense and research-intensive paper into something once removed from heavier academic barriers for an audience that might not have a technical background. The goal is to just tell the story of the project concisely and powerfully, to highlight the important findings without getting into the nitty-gritty of the challenges. I have a week or so to turn over structure, visualizations of the case study bridges (the Third Street Bridge and the Jackson Avenue span), so they feel real and relatable to anyone in the room.
Week 12: Final week
Hi everyone! This past Friday, June 12th, I gave my final senior project presentation! I presented to a really lovely audience, and it was great to finally put everything together into one cohesive story: the research, the case studies, the framework, all of it. Looking back at where this project started months ago versus where it ended up, it's great to see what it has amounted to.
I want to say a huge thank you to everyone who made this possible: Ms. Olender, my faculty advisor, for guiding me through this whole process; Zhihua Yi, my on-site mentor at the DOT, for all the time and insight; Farid from the DOT for helping with the logistics; and DOT Commissioner Mike Flynn for their support along the way; and BASIS for giving me the opportunity to do a project like this in the first place. I'm really grateful for everything I learned, and I'm excited to bring this curiosity about infrastructure, data, and AI with me into college and beyond.
Here is a link to my presentation if interested:
https://docs.google.com/presentation/d/1gXBFIQO2Dk2nEklcy78lXYZgLZa-JZNYSytL5RXAafw/edit?usp=sharing