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.
翻译:消防部门每24秒钟就要求消防部门作出反应。许多消防员在搜索和救援行动中因隐蔽的危险而受伤。此外,防火水径流污染可能威胁人体健康。我的目标是开发一个系统,监测灾害引起的环境损害,向事件指挥人员提供早期情报。我开发了一个多光谱遥感系统,检查空气和水质质量,以开展更安全和无障碍的有害环境作业。关键组成部分包括:(a)配备4个传感器的无人驾驶飞机(气体传感器、热相机、全球定位系统传感器、视像相机)和无线通信器,供检查之用,(b)AI型计算机视像基地站,以确定目标,(c)低成本、便携式、光谱水质分析器和(d)机械回收器。原型显示为消防员和科学家提供更安全、更方便的搜索和救援行动的潜力。气体传感器可以识别厚烟雾状况(临界值 > 400)。视觉和热摄像头探测隐藏的热物体,并将图像发送给AI-动力分析仪,以便识别和稳定目标质量。水质量,通过全球定位系统对目标进行检索。