https://scholars.lib.ntu.edu.tw/handle/123456789/630483
標題: | Unpacking data-centric geotechnics | 作者: | Phoon, Kok Kwang JIAN-YE CHING Cao, Zijun |
關鍵字: | Bayesian machine learning | Data-centric geotechnics | Data-driven site characterization (DDSC) | Data-informed decision support index | Project DeepGeo | 公開日期: | 1-十二月-2022 | 出版社: | KEAI PUBLISHING LTD | 卷: | 7 | 期: | 6 | 起(迄)頁: | 967 | 來源出版物: | Underground Space (China) | 摘要: | The purpose of this paper (presented online as a keynote lecture at the 25th Annual Indonesian Geotechnical Conference on 10 Nov 2021) is to broadly conceptualize the agenda for data-centric geotechnics, an emerging field that attempts to prepare geotechnical engineering for digital transformation. The agenda must include (1) development of methods that make sense of all real-world data (not selective input data for a physical model), (2) offering insights of significant value to critical real-world decisions for current or future practice (not decisions for an ideal world or decisions of minor concern to geotechnical engineers), and (3) sensitivity to the physical context of geotechnics (not abstract data-driven analysis connected to geotechnics in a peripheral way, i.e., engagement with the knowledge and experience base should be substantial). These three elements are termed “data centricity”, “fit for (and transform) practice”, and “geotechnical context” in the agenda. Given that a knowledge of the site is central to any geotechnical engineering project, data-driven site characterization (DDSC) must constitute one key application domain in data-centric geotechnics, although other infrastructure lifecycle phases such as project conceptualization, design, construction, operation, and decommission/reuse would benefit from data-informed decision support as well. One part of DDSC that addresses numerical soil data in a site investigation report and soil property databases is pursued under Project DeepGeo. In principle, the source of data can also go beyond site investigation, and the type of data can go beyond numbers, such as categorical data, text, audios, images, videos, and expert opinion. The purpose of Project DeepGeo is to produce a 3D stratigraphic map of the subsurface volume below a full-scale project site and to estimate relevant engineering properties at each spatial point based on actual site investigation data and other relevant Big Indirect Data (BID). Uncertainty quantification is necessary, as current real-world data is insufficient, incomplete, and/or not directly relevant to construct a deterministic map. The value of a deterministic map for decision support is debatable. The computational cost to do this for a 3D true scale subsurface volume must be reasonable. Ultimately, geotechnical structures need to be a part of a completely smart infrastructure that fits the circular economy and need to focus on delivering service to end-users and the community from project conceptualization to decommission/reuse with full integration to smart city and smart society. Although current geotechnical practice has been very successful in taking “calculated risk” informed by limited data, imperfect theories, prototype testing, observations, among others and exercising judicious caution and engineering judgment, there is no clear pathway forward to leverage on big data and digital technologies such as machine learning, BIM, and digital twin to meet more challenging needs such as sustainability and resilience engineering. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/630483 | ISSN: | 20962754 | DOI: | 10.1016/j.undsp.2022.04.001 |
顯示於: | 土木工程學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。