Operational-Level Optimization of Autonomous Logistics Enterprises Transition of Research Efforts to the University of Alabama
Abstract
The purpose of this supplemental document is to support the transition of research to the University of Alabama (UA). The PI, Kevin M. Curtin, PhD has taken a position as Professor of Geography at UA, and both GMU and UA have stated their position that the research must transition to UA in order to continue to successful completion. The overall goals and deliverables from the original proposal have not changed, and will not change as a result of this transition. The motivational problem statement for this research continues to be that, as logistics systems move from being unmanned and remotely-piloted to being autonomous and supervised there is greater need to use high-fidelity models and to develop algorithmic approaches to enhance precision and reduce risk in logistics operations. More specifically, this research examines the “platform mix problem”, asking “How does an organization efficiently allocate the mixed set of autonomous logistics systems needed to support distributed operations?” The research being undertaken on this project directly addresses this problem through the Autonomous Logistics Operations Family of Tools (ALOFT). Significant progress has been made to date in the areas of scenario development, formulation of the overall logistics delivery problem with alternate objective functions, and presentation of Pareto Trade-off Analysis to permit platform mix decision-making. The core of the approach to evaluate autonomous platform mix rests in the integration of optimization methods with Geographic Information Systems technologies and spatial analytical techniques. Tools for such comparisons using Pareto trade-off boundaries are under continuing development. With these tools decision-makers are able to examine the tradeoffs in cost and performance and make determinations as to the importance of each and the procurement or development efforts that should be undertaken. More broadly a set of functionality has been combined into a Testbed Environment in order to implement the scenarios and the optimization approach described above. This Testbed Environment consists of a set of tightly integrated tools with off-the-shelf GIS functionality, custom GIS scripting, a linkage to linear programming solution software, and customized display tools. The Testbed Environment supports scenario development, database storage and maintenance, data generation through GIS analysis, solution of the optimization problem instances, and storage and display of the results of those model runs. We view the Testbed Environment as a value-added deliverable for the project, and as a potential link to further applied research that can potentially transition the results of the ALOFT project into an operationalized planning system. Continued work will include building a library of testing scenarios, and additional optimization formulations and constraint sets. The initial results to date support the assertion that there are significant basic scientific contributions of the research being undertaken, including a successful exploration of the operational level characteristics of logistics systems in a new problem domain, and the development of efficient computational approaches to solving autonomous logistics problems leading to the evaluation of platform mix. Moreover, we continue to assert that solving the platform mix problem is relevant to Naval operations because rigorous answers will enable policy makers, program managers, and engineers to confidently invest time, effort, and resources in the combination of systems that are most likely to be effective and useful when the revolution in autonomous logistics actually occurs.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 23, 2018
- Source ID
- N000141812094
Entities
People
- Kevin M. Curtin
Organizations
- Office of Naval Research
- United States Navy
- University of Alabama