从确定性假设到不确定性感知：数据驱动的工程决策新范式

Modern civil engineering faces increasing demands for safety, sustainability, and resilience. To meet these challenges, engineers must make informed decisions under uncertainty, leveraging both predictive models and data. While traditional engineering approaches have focused on deterministic modeling, the engineering community is starting to realize the importance of explicitly incorporate uncertainty and use data to guide decisions throughout the lifecycle of infrastructure systems.

In this talk, Professor Straub will present how formal decision analysis can enhance the use of engineering predictions in practice and support data-informed decision making. Professor Straub will emphasize the importance of explicitly addressing uncertainty and the sequential nature of decision-making, e.g., by means of probabilistic digital twins. Sensitivity analysis will be discussed as a key tool for understanding model behavior and communicating limitations. These concepts will be illustrated through practical applications, including flood risk planning, construction, infrastructure development and predictive maintenance strategies.

      Daniel Straub是慕尼黑工业大学（TUM）工程风险分析讲席教授。其研究专注于基于物理和数据驱动的考虑不确定性的工程序贯决策建模与分析。其于2000年和2004年在苏黎世瑞士联邦理工大学先后获得土木工程学位及理学博士学位。之后，其赴加州大学伯克利分校担任研究和教学职位，并于2008年加入慕尼黑工业大学接替退休的被美国土木工程学会评价为“40多年来一直是研究考虑不确定性下结构行为理论的国际领导者“的Rüdiger Rackwitz教授获得讲席。Straub教授也活跃于工程咨询，曾先后于瑞士和德国等地创办和参与多家工程咨询公司。其在领域内的贡献广受认可，获得诸多国际学会颁发的学术奖励。其于2023年获得由土木工程风险与可靠性协会（CERRA）每四年颁发一次的C. Allen Cornell奖，该奖项旨在表彰在土木工程风险与可靠性理论的科学研究和应用方面取得的杰出成就。其目前是《ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems》期刊的副主编，《Structural Safety》、《Reliability Engineering & System Safety》和《Probabilistic Engineering Mechanics》等多个期刊的编委。