一个早期预警的AI驱动的便携式系统，以减少工作量和检查自然灾害后的环境破坏
An early warning AI-powered portable system to reduce workload and inspect environmental damage after natural disasters
摘要：
1.3数百万家庭火灾，3,400名平民死亡，230亿美元的损失，消防部门每24秒就被要求作出反应。多名消防员因隐患在搜救行动中受伤。此外，阻燃剂水径流污染可威胁人类健康。我的目标是开发一个系统来监测灾害引起的环境破坏，为事故指挥官提供早期情报。我开发了一种多光谱传感系统，用于检测空气和水质，以实现更安全和可访问的危险环境操作。关键部件包括a）安装有四个传感器（气体传感器、热成像摄像机、GPS传感器、视觉摄像机）和无线通信器的无人机，b）用于识别目标的人工智能计算机视觉基站，c）低成本、便携式光谱水质分析仪和d）机器人回收器。该原型展示了为消防员和科学家提供更安全和更方便的搜索和救援行动的潜力。气体传感器可以识别浓烟情况（阈值> 400）。视觉和热成像摄像头检测到隐藏的热物体，并将图像发送到人工智能驱动的分析仪，以识别和定位目标，并使用救援GPS坐标进行机器人检索。用光谱特征分析水质，以指示与潜在污染物相关的浊度水平（阈值> 1.3）。原型结果已向森尼韦尔消防部门展示，并收到了令人鼓舞的反馈。未来的目标包括用智能服装监测消防员的健康和过度劳累。
英文摘要：
1.3 million household fires, 3,400 civilian deaths, and 23 billion dollars in damage, a fire department is called to respond every 24 seconds. Many firefighters are injured during search and rescue operations due to hidden dangers. Additionally, fire-retardant water runoff pollution can threaten human health. My goal is to develop a system to monitor calamity-induced environment damage to provide early-intelligence to incident-commanders. I have developed a multi-spectral sensing system to inspect air and water quality for safer and accessible hazardous environment operations. Key components include a) drone mounted with four sensors (gas sensors, thermal camera, GPS sensor, visual camera) and wireless communicator for inspection, b) AI-powered computer vision base-station to identify targets, c) low-cost, portable, spectral water quality analyzer and d) robotic retriever. The prototype demonstrates the potential for safer and more accessible search and rescue operations for fire-fighters and scientists. The gas sensor could identify thick smoke situations (thresholds > 400). The visual and thermal cameras detected hidden hot objects and sent images to AI-powered analyzer to identify and localize target with rescue GPS coordinates for robotic retrieval. Water quality was analyzed with spectral signatures to indicate turbidity levels that correlate with potential pollutants (threshold > 1.3). Prototype results were shown to the Sunnyvale fire department and received encouraging feedback. Future goals include monitoring firefighter health and overexertion with smart clothes.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
