CAREER: Advanced demand estimators for energy-efficiency in personal transportation

1253475 (Daziano)。PI的长期职业目标是为决策分析的研究和教育做出重大贡献，以更好地了解消费者对节能工程技术的行为反应。为了实现这一目标，CAREER的研究目标是利用微观计量离散选择理论更好地为低排放车辆的工程设计提供信息。旨在培养工程师创造社会愿意采用的技术上合理的解决方案的长期目标，CAREER教育目标是为未来的工程师提供基于消费者需求的工程决策的多学科视野。本提案的研究计划旨在：1)对具有异质消费者和节能汽车技术决策规则的大型联立方程离散选择系统的结构参数生成新的需求估计量；2)推导出考虑行为不确定性的支付意愿和消费者剩余测度的非参数贝叶斯估计量；3)对低排放汽车部署的综合反事实情景进行系统的贝叶斯成本效益分析，为政策、技术、工程和基础设施规划决策提供信息。这三项研究任务将使用消费者采用不同来源的低排放车辆的数据进行验证和测试。教育计划建立在计划研究的基础上，并将有助于计划研究，包括四个步骤：1)与康奈尔大学的可持续发展研究中心合作，为不同背景但对能源可持续发展有共同兴趣的大学生提供多学科研究和学习经验；2)广泛推广可持续旅行行为，以教育公众，并激励社会多元化的后代追求工程事业；3)通过实施和不断改进旨在将需求侧动力学融入工程的三门课程，加强高级本科和研究生阶段的课程设置；4)指导博士和孟阶段的研究生。主要研究成果将是基于随机效用最大化的复杂结构方程系统的联合估计问题的解决方案，该问题可用于制定能源效率的需求模型。未解决的计量经济学挑战将解决为确定的简化参数导出灵活的贝叶斯参数和非参数模拟辅助推理。联立方程需求模型将考虑消费者反应与经济、环境、能源和运输系统的相互作用。需求模型还将纳入社会象征价值，如亲环境偏好、能源安全问题，以及消费者对新兴可持续技术的认识和他们采用这些技术的意愿。为了表示不断发展的技术之间的选择，需求系统还将考虑能源多样化、选择动态和定性属性的测量。通过与福特汽车、加州大学伯克利分校、欧洲经济研究中心和罗马大学的合作，将促进对数据的访问。这些技术成果将有助于需要在不确定情况下进行决策的领域。从需求估算器中得出的决策分析工具不仅将用于评估先进节能推进技术和基础设施的定价和投资策略，还将用于评估公共政策和激励措施，以最好地促进工业向低排放车辆的转变和消费者对低排放车辆的接受。研究结果不仅对美国的政策制定者和交通规划者意义重大，而且对汽车制造商改进汽车工程技术也具有重要意义。知识转移将通过康奈尔大学内外的多学科合作实现，包括国际合作伙伴关系。教育活动将侧重于传播消费者对成功工程解决办法的反应的相关性。康奈尔大学工程学院的教学卓越研究所将协助使用个人反应系统创建创新的主动选择实验。公开讨论放映关于电动汽车的纪录片，以及每年参加纽约车展，将支持向青年和大学新生推广可持续旅行行为的计划，特别关注代表性不足的群体。

1253475 (Daziano). The long-term career goal of the PI is to contribute significantly to both research and education in decision-making analysis to better understand consumer behavioral response to energy-efficient engineered technologies. In pursuit of this goal, the CAREER research objective is to exploit microeconometric discrete choice theory to better inform engineering of low emission vehicles. Aiming at the long-term goal of preparing engineers to create technically sound solutions that society is willing to adopt, the CAREER educational objective is to provide future engineers with a multidisciplinary vision of engineering decision-making informed by consumer demand. The research plan of this proposal seeks to: 1) generate new demand estimators of the structural parameters of a large-scale simultaneous-equations discrete choice system with heterogeneous consumers and decision rules for energy-efficient automobile technologies, 2) derive nonparametric Bayesian estimators of willingness-to-pay and consumer-surplus measures that account for behavioral uncertainties, 3) formulate a systematic Bayesian cost-benefit analysis of integrative counterfactual scenarios of low emission vehicle deployment for informing policy, technology, engineering, and infrastructure planning decisions. The three research tasks will be validated and tested using data on consumer adoption of low-emission vehicles from different sources. The education plan builds on and will contribute to the planned research, and comprises four steps: 1) collaboration with sustainability research centers at Cornell to foster multidisciplinary research and learning experiences for college students with different backgrounds but with common interests in energy sustainability, 2) extensive outreach on sustainable travel behavior to educate the public and to motivate socially-diverse future generations to pursue engineering careers, 3) enhancement of the curriculum at the senior undergraduate and graduate levels through implementing and continuously improving three courses aimed at integrating demand-side dynamics into engineering, and 4) mentoring graduate students at the PhD and MEng levels. The main research outcome will be a solution for the joint estimation problem of a complex system of structural equations based on random utility maximization that can be applied to formulate demand models for energy efficiency. Unsolved econometric challenges will be addressed for deriving flexible Bayesian parametric and nonparametric simulation-aided inference for identified reduced parameters. The simultaneous equations demand model will account for interactions of consumer response with the economic, environmental, energy, and transportation systems. The demand model will also incorporate social symbolic values such as pro-environmental preferences, energy security concerns, as well as consumers' awareness of emerging sustainable technologies and their readiness to adopt them. To represent choice among continuously evolving technologies, the demand system will also account for energy diversification, choice dynamics, and measurement of qualitative attributes. Access to data will be facilitated through collaborations with Ford Motors, UC Berkeley, the Centre for European Economic Research, and the University of Rome 3. The technical results will contribute to fields where decision-making under uncertainty is needed. Decision-making analysis tools derived from the demand estimators will serve to evaluate not only pricing and investment strategies for advanced energy-efficient propulsion technologies and infrastructure, but also public policies and incentives to best promote industry conversion to and consumer acceptance of low-emission vehicles. The results will be significant not only for US policymakers and transportation planners, but also for informing auto manufacturers to improve how industry engineers vehicles. Knowledge transfer will be achieved through multidisciplinary collaboration within and beyond Cornell, including international partnerships. Educational initiatives will be focused on disseminating the relevance of consumer response for successful engineering solutions. Cornell Engineering's Teaching Excellence Institute will assist in creating innovative active choice experiments using personal-response systems. Publicly discussed screenings of documentaries about electric vehicles and yearly participation at the NYS fair will support outreach plans on sustainable travel behavior to youth and college freshmen, with a special focus on underrepresented groups.