PFI-RP: Data-Driven Services for High Performance and Sustainable Buildings

PFI-RP：面向高性能和可持续建筑的数据驱动服务

• 批准号: 1827757
• 负责人: Zheng O'Neill
• 金额: $ 75万
• 依托单位: University of Alabama Tuscaloosa
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-09-15 至 2020-10-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1827757&HistoricalAwards=false
• 关键词: PFI; RP; Data; Driven; Services

The broader impact/commercial potential of this PFI project will lead to the creation of a truly new breed of building services companies that can provide guaranteed building performance. The results from this proposal will contribute to both building energy efficiency research and practices. Beyond energy conservation, outcomes of the project also have an important societal benefit in advancing the role of a skilled workforce in conjunction with automation. In addition, the proposed effort builds an innovation ecosystem which spans the supply chain for building services and building analytics and automated intelligence suppliers. The proposed research will fulfill an important role in sustainable environments and will enable services with significant economic and human impact. While the building service industry is the focus here, the results obtained have the potential to lead to a better understanding of other human-centered services driven by big data, and will influence the development of novel service platforms applicable to urban operations, such as city-wide transportation control, power grids, and public health, in which system-wise fault discovery and recovery have to be achieved through dynamic spatio-temporal data streams of varying quality and coverage.The proposed project of Building Doctor's Medicine Cabinet (BDMC) service platform, empowered by novel multi-resolution (temporal and spatial) data analytics, high-dimensional robust modeling, and human-centered interface design, will help building doctors, who are engineers and technicians working in the building service industry, to effectively troubleshoot building problems and to identify systematic and prognostic solutions. BDMC will integrate building datasets from in-situ and control system measurements, knowledge from building doctors, and leverage the patterns and anomalies discovered on this data to (1) diagnose and prognose whole building problems with greatly reduced engineering labor input, false alarms and false dismissals, 2) identify building system hierarchy and develop data-driven energy models; and (3) provide human-centered data visualization and feedback (fault impact analysis and prioritization). The proposed effort will transform the current labor-intensive building service industry into a smart service industry. The main outcomes from this PFI-RP project are algorithms and codes that can be licensed to the industry to be integrated with their existing market products.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

这个PFI项目的更广泛的影响/商业潜力将导致创建一个真正的新型建筑服务公司，可以提供有保证的建筑性能。该方案的研究成果将有助于建筑节能研究和实践。除了节能之外，该项目的成果还在促进熟练劳动力与自动化的结合方面具有重要的社会效益。 此外，拟议的努力建立了一个创新生态系统，该生态系统跨越了建筑服务、建筑分析和自动化智能供应商的供应链。拟议的研究将在可持续环境中发挥重要作用，并将使服务具有重大的经济和人类影响。虽然建筑服务行业是这里的重点，但所获得的结果有可能导致更好地理解由大数据驱动的其他以人为本的服务，并将影响适用于城市运营的新型服务平台的开发，例如城市交通控制，电网和公共卫生，其中，系统级故障发现和恢复必须通过动态时空数据流来实现，这些数据流具有不同的质量和覆盖范围。通过新颖的多分辨率（时间和空间）数据分析，高维鲁棒建模和以人为中心的界面设计，将帮助建筑医生，他们是在建筑服务行业工作的工程师和技术人员，有效地解决建筑问题，并确定系统和预测解决方案。BDMC将整合来自现场和控制系统测量的建筑数据集，建筑医生的知识，并利用这些数据中发现的模式和异常来（1）诊断和排除整个建筑问题，大大减少工程劳动力投入，错误警报和错误解雇，2）识别建筑系统层次结构并开发数据驱动的能源模型;以及（3）提供以人为中心的数据可视化和反馈（故障影响分析和优先级排序）。 这项计划将把目前劳动密集型的建筑服务业转变为智能服务业。该PFI-RP项目的主要成果是可以授权给业界与其现有市场产品集成的算法和代码。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Acquisition of Typical Occupancy Schedules for Commercial Buildings from Social Networks

从社交网络获取商业建筑的典型入住表

• DOI: 10.1145/3363459.3363537
• 发表时间: 2019
• 期刊: and Visualization
• 作者: Lu, Xing;Feng, Fan;O'Neill, Zheng
• 通讯作者: O'Neill, Zheng

CA-Smooth: Content Adaptive Smoothing of Time Series Leveraging Locally Salient Temporal Features

CA-Smooth：利用局部显着时间特征的时间序列的内容自适应平滑

• DOI: 10.1145/3297662.3365830
• 发表时间: 2019
• 期刊: the 11th International Conference on Management of Digital EcoSystems
• 作者: Rossini, Rosaria;Poccia, Silvestro;Candan, K. Selcuk;Sapino, Maria Luisa
• 通讯作者: Sapino, Maria Luisa

Generalized Ordinal Learning Framework (GOLF) for Decision Making with Future Simulated Data

用于利用未来模拟数据进行决策的广义序数学习框架 (GOLF)

• DOI: 10.1142/s0217595919400116
• 发表时间: 2019
• 期刊: Asia-Pacific Journal of Operational Research
• 影响因子: 1.4
• 作者: Pedrielli, Giulia;Selcuk Candan, K.;Chen, Xilun;Mathesen, Logan;Inanalouganji, Alireza;Xu, Jie;Chen, Chun-Hung;Lee, Loo Hay
• 通讯作者: Lee, Loo Hay

Extracting typical occupancy schedules from social media (TOSSM) and its integration with building energy modeling

• DOI: 10.1007/s12273-020-0637-y
• 发表时间: 2020-05-13
• 期刊: BUILDING SIMULATION
• 影响因子: 5.5
• 作者: Lu, Xing;Feng, Fan;O'Neill, Zheng
• 通讯作者: O'Neill, Zheng

SAN : Scale-Space Attention Networks

• DOI: 10.1109/icde48307.2020.00079
• 发表时间: 2020-04
• 期刊: 2020 IEEE 36th International Conference on Data Engineering (ICDE)
• 作者: Yash Garg;K. Candan;M. Sapino