Human and modeling approaches for humanitarian transportation planning
Author(s)Gralla, Erica Lynn
Massachusetts Institute of Technology. Engineering Systems Division.
MetadataShow full item record
Recent disasters have highlighted the need for more effective supply chain management during emergency response. Planning and prioritizing the use of trucks and helicopters to transport humanitarian aid to affected communities is a key logistics challenge. This dissertation explores ways to improve humanitarian transportation planning by building on the strengths of both humans and models. The changing, urgent, multi-objective context of humanitarian aid makes it challenging to formulate and deploy useful planning models. Humans are better able to understand the context, but struggle with the complexity of the problem. This research investigates the strengths and weaknesses of human transportation planners in comparison with models, with the goal of supporting both- better human decision-making and better models for humanitarian transportation planning. Chapter 2 investigates how experienced humanitarian logisticians build transportation plans in a simulated emergency response. Based on an ethnographic study of ten logistics response teams, I show how humans come to understand the problem and its objectives through sensemaking, and solve it through a search-like series of decisions guided by goal-oriented decision rules. I find that the definition of objectives is an important strength of the sensemaking process, and that the human reliance on greedy search may be a weakness of human problem-solving. Chapter 3 defines a performance measure for humanitarian transportation plans, by measuring the importance of the objectives identified in the ethnographic study. I use a conjoint analysis survey of expert humanitarian logisticians to quantify the importance of each objective and develop a utility function to value the performance of aid delivery plans. The results show that the amount of cargo delivered is the most important objective and cost the least; experts prefer to prioritize vulnerable communities and critical commodities, but not to the exclusion of others. Chapter 4 investigates the performance of human decision-making approaches in comparison to optimization models. The human decision-making processes found in Chapter 2 are modeled as heuristic algorithms and compared to a mixed-integer linear program. Results show that optimization models create better transportation plans, but that human decision processes could be nearly as effective if implemented consistently with the right decision rules.
Thesis (Ph. D.)--Massachusetts Institute of Technology, Engineering Systems Division, 2012.Cataloged from PDF version of thesis.Includes bibliographical references.
DepartmentMassachusetts Institute of Technology. Engineering Systems Division.
Massachusetts Institute of Technology
Engineering Systems Division.