PFI-TT: Prototyping a Hybrid System for Automated Activity Detection and Monitoring of Construction Equipment

这种创新技术转化伙伴关系（PFI-TT）的更广泛影响/商业潜力在于解决建筑行业缺乏成本效益高且易于使用的系统来识别和监控工地作业的问题。拟议的系统是一种新颖的解决方案，有望帮助现场管理人员和设备所有者更好地评估设备的工作状态，评估安全条件，并在必要时采取预防/纠正措施。该系统自动整合和处理从多台机器收集的数据的能力，对于中型和大型公司来说尤其有益，这些公司利用施工设备车队来执行大型项目。该技术开发项目将推进“自动化施工绩效监控”的概念，从简单的“位置跟踪”工具，发展为更有信息和全面的“行动跟踪和控制”系统。研究表明，用于建筑和采矿的重型商用车约占北美车队管理系统总市场规模的6%（4200万美元）。本文的研究重点是车队管理系统的自动跟踪和监控部分。拟议的项目旨在创建和经验测试一个混合系统的可行性，该系统能够识别、跟踪和记录工地施工设备进行的活动，然后处理和整合从单个机器捕获的数据，以自动生成和更新项目（和公司）级别的管理信息。该系统由硬件和数据分析两部分组成。在本项目中专门设计和制造的硬件单元将安装在每一件施工设备上。该硬件单元能够捕获机器在执行不同任务时产生的三种主要数据源（音频、运动和位置数据）。硬件单元还能够将收集的数据无线传输到位于任何物理或虚拟位置（如项目办公室）的中央处理单元。对每台机器收集的数据进行处理，并根据声音和运动模式识别机器的不同活动。活动的时间顺序列表将被转换为关于生产率和完成工作数量的语义信息。为单个机器收集和计算的信息将在项目（和公司）级别进行集成和处理。该奖项反映了美国国家科学基金会的法定使命，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。

The broader impact/commercial potential of this Partnerships for Innovation - Technology Translation (PFI-TT) is in addressing the lack of cost-effective and easy-to-use systems for recognizing and monitoring jobsite operations in the construction industry. The proposed system is a novel solution expected to assist jobsite managers and equipment owners to better assess the working status of equipment, to evaluate safety conditions and take preventive/corrective measures, when necessary. The ability of the system in automatically integrating and processing data collected from multiple machines will be particularly beneficial for medium and large size companies where fleets of construction equipment are utilized to execute larger size projects. This technology development project will advance the concept of “automated construction performance monitoring” from simple “location tracking” tools into more informative and comprehensive “action tracking and control” systems. Research shows that heavy commercial vehicles used in construction and mining represent approximately 6% ($42 million) of the total North America fleet management system market size. The proposed research focuses on the automated tracking and monitoring component of the fleet management system. The proposed project intends to create and empirically test the feasibility of a hybrid system capable of recognizing, tracking, and recording activities conducted by construction equipment at jobsites and then process and integrate the captured data from individual machines to automatically generate and update managerial information at project (and company) levels. The system consists of a hardware unit and the data analysis. The hardware unit, specifically designed and fabricated during this project, will be mounted on each piece of construction equipment. This hardware unit is capable of capturing three major sources of data (audio, kinematic, and location data) produced by machines while performing different tasks. The hardware unit is also capable of wirelessly communicating the collected data into a central processing unit located at any physical or virtual location such as a project office. The collected data for each piece of machinery will be processed and the different activities of the machine will be recognized based on the sounds and kinematic patterns. The chronological list of activities will be converted into semantic information regarding productivity rates and quantities of completed work. The information collected and computed for individual machines will be integrated and processed at the project (and company) level.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.