NuRail Project ID | NURail2016-UIUC-R17 |
Project Title | Optimal Planning of Rail Grinding Activities in Large-scale Networks |
University | University of Illinois at Urbana-Champaign |
Project Manager | Yanfeng Ouyang |
Principal Investigator | Yanfeng Ouyang |
PI Contact Information | |
Funding Source(s) and Amounts Provided (by each agency or organization) | $65,000 NURail, $32,500 CSX |
Total Project Cost | $97,500 |
Agency ID or Contract Number | DTRT13-G-UTC52 (Grant 2) |
Start Date | 2016-08-16 |
End Date | 2017-08-15 |
Location | |
Brief Description of Research Project | Every year, U.S. railroads spend millions of dollars on grinding of rails so as to prevent wear and fatigue of rail steel from proceeding at an increasing pace. Usually, a fleet of rail-bound grinders (and the associated crew) must travel across in a large-scale railroad network to restore the condition of rail tracks. Thousands of track segments must be grinded at certain frequencies (varying from once a few months to once a couple of years) to ensure safety of train operations, and the schedule of the rail grinders must typically be updated every a few months. The schedule should not only meet the grinding frequency requirement of each segment but also satisfy complex interrelations among these segments (e.g., geography- and traffic-based priorities, mutually exclusiveness, and time windows). These constraints could be either hard (i.e., should not be violated at all) or soft (i.e., may be violated with a penalty). Sometimes, resource planning is also needed in order to determine the optimal grinder fleet size and to balance workload throughout the years. This project creates a decision-support system (including advanced mathematical models and efficient solution algorithms) for the optimal scheduling of rail grinding activities and economic planning of resources, which can (i) enhance railroad safety by better satisfying the grinding frequency requirements, and (ii) help reduce operation costs by minimizing the number of grinders, travel costs, overtime work hours, and resource conflicts. |
Describe Implementation of Research Outcomes (or why not implemented) | The project intends to develop advanced mathematical models and solution techniques for the rail grinding scheduling problem. The developed models and tools will be incorporated into a decision-support system that can be used to plan resources and optimize scheduling of rail grinding activities in large scale railroad networks. The following tasks are going to be carried out throughout this project: 1. Conduct intensive literature review to understand the current practice on rail grinding scheduling domestically and internationally, in order to identify the state of art and research gap. 2. Develop mathematical formulations and solution algorithms for the grinding scheduling problem, taking into consideration various real-world constraints and business rules. An analytical approach will be used. We will build upon existing literature on multi-commodity network flows and network routing to develop optimization models. Mathematical programming and systematic optimization methods (e.g., relaxation and/or resource-direct decomposition, and meta-heuristics) will be developed to find the best system-wide schedule that satisfies all time-window and priority requirements. The key step toward formulating the problem is the construction of a network with novel definitions of nodes, arcs, costs and capacities such that the minimum cost flow solution can be mapped into the optimal grinding schedule. 3. Since the scheduling problem will be raised every a few months at various spatial levels, we aim to automate the decision-support procedure for a general railroad network and an arbitrary planning horizon. Our mathematical formulation and optimization techniques will be coded into a computer program that can be used directly for real-world applications. 4. Conduct technology transfer and field implementation of the proposed models using real-world data from a Class-1 railroad company. This company is providing the required cost-share for this project. 5. Identify research needs and future research directions. 6. Document and publish the research results in academic journals and conference presentations. The final deliverable will be 1-2 journal papers documenting the research work. A graduate research assistant and a part-time post-doc research associate will be working on the project. |
Photos | |
Impacts/Benefits of Implementation (actual, not anticipated) | This project addresses at least two of the U.S. Department of Transportation (DOT)�s strategic goals: Safety, and State of Good Repair, because properly planned rail grinding activities remove rail (and wheel) surface deformation and cracks, improve the dynamic stability and safety of rolling stock, and elongate overall infrastructure life. This project spans a wide range of NURail topic areas, such as Infrastructure, Safety & Risk, Planning, Freight, and Technology Transfer. The project will advance our basic understanding on how to most efficiently and effectively utilize resources and maintain conditions of railroad infrastructures, thereby providing guidance on tactical and strategic planning, operational management, and public and private sectors� policy making. The planned technology transfer efforts and ongoing collaboration with a Class-1 railroad company will ensure relevance and applicability of the research outcome. |
Web Links | |
Reports | |
Project Website | |
Completed | yes |
Final Report | NURail2016-UIUC-R17_Final_Report.pdf |