CAREER: Less Obstructive Vital Evaluation-inspection Robots for Bridges

桥梁是重要的基础设施组成部分，对公众出行安全和经济的可持续性至关重要。美国60多万座桥梁的平均年龄为42年，由于机械磨损和天气条件、维护不足以及检查和评估不足，桥梁受到了不同程度的恶化。检查、修理和更换老化的公路桥梁的费用估计超过1400亿美元。桥梁坍塌——包括华盛顿州的I-5斯卡吉特河大桥和密西西比河上的I-35W大桥，造成13人死亡，超过145人受伤——应该成为确保桥梁状况得到准确评估的强大动力。非破坏性评估（NDE）技术提供了一种识别和预测桥梁早期退化的方法，从而实现主动修复和修复。然而，目前这些技术依靠人工采集数据，容易出现人为错误，且不全面、速度慢、劳动密集、成本高，对交通流量产生负面影响，无法满足日益增长的高效、经济、安全的桥梁检测需求。更重要的是，人工检查对检查员来说是危险的，因为他们必须在开放的交通中工作，并且要爬上很高的桥梁结构。为了解决这些问题，该项目试图将NDE技术整合到一个系统中，该系统将允许机器人团队对桥梁进行检查。除了提高人类桥梁检查员的安全性，减少交通延误，并通过减少人工劳动节省资金外，该系统还将结合每个NDE机器人收集的数据，对桥梁状况进行更准确、更可靠的评估。该项目的目标是开发一个新的研究框架，将NDE技术与机器人技术和自动化科学的最新进展相结合。具体而言，本项目的目标是：(1)为多个无损检测传感器融合建立一个通用的理论框架，并证明其可行性，以促进高效、准确和可靠的公路桥梁结构检测；(2)开发一种控制框架，以协调具有NDE传感器融合和人在环的多个检测机器人，实现安全、准确和高效的桥梁检测，同时最大限度地减少交通中断和劳动密集型操作；(3)在实验室环境和实际桥梁上对框架系统进行测试和评估。该项目的长期研究目标是将充分体贴的机器人智能从理论到实际应用，如民用基础设施检查，以挽救生命，降低检查和维护成本。该项目由计算机和信息科学与工程理事会（CISE）的信息和智能系统（IIS）部门以及刺激竞争研究的既定计划（EPSCoR）共同资助。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。

Bridges are essential infrastructure components that are critical for the safety of the traveling public and the sustainability of the economy. The more than 600,000 bridges in the United States have an average age of 42 years and have been subjected to varying levels of deterioration due to mechanical wear and weather conditions, inadequate maintenance, and deficiencies in inspection and evaluation. The cost of inspecting, repairing, and replacing deteriorating highway bridges has been estimated at more than $140 billion. Bridge collapses - including the I-5 Skagit River Bridge in Washington state and the I-35W bridge over the Mississippi River, which killed 13 people and injured more than 145 - should serve as strong motivation to ensure that the conditions of bridges are accurately assessed. Non-destructive evaluation (NDE) technologies present a way to identify and predict early-stage bridge deterioration to enable proactive repair and rehabilitation. However, these technologies in their current state cannot meet the increasing demand for efficient, cost-effective, safe bridge inspection because they rely on manual data collection, which is prone to human error, non-comprehensive, slow, labor-intensive and therefore expensive, and negatively impacts traffic flow. More importantly, manual inspection is dangerous for inspectors because they must work in open traffic and climb high bridge structures. In response to these concerns, this project seeks to integrate NDE technologies into a single system that will allow bridges to be inspected by a team of robots. In addition to increasing the safety of human bridge inspectors, reducing traffic delays, and saving money by reducing manual labor, this system will combine the data collected by each individual NDE robot to build a more accurate and reliable assessment of a bridge's condition.The goal of this project is to develop a novel research framework that will integrate NDE technologies with recent advances in robotics and automation sciences. Specifically, the objectives of this project are: (1) to develop a general theoretical framework for multiple NDE sensor fusion and to demonstrate its feasibility to facilitate efficient, accurate, and reliable inspection of highway bridge structures; (2) to develop a control framework to coordinate multiple inspection robots with NDE sensor fusion and with human-in-the-loop for safe, accurate, and efficient bridge inspection while minimizing traffic interruption and labor-intensive operations; and (3) to test and evaluate the framework system both in the lab environment and on real bridges. The long-term research goal of this project is to bring full considerate robot intelligence from theory to real-world applications, such as civil infrastructure inspection, in order to save lives and reduce the costs of inspection and maintenance.This project is jointly funded by the Information and Intelligent Systems (IIS) Division in the Computer and Information Science and Engineering (CISE) Directorate and the Established Program to Stimulate Competitive Research (EPSCoR).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.