NRI: FND: COLLAB: Distributed Bayesian Learning and Safe Control for Autonomous Wildfire Detection

野火摧毁了数百万公顷的森林、敏感的生态系统和人类基础设施。减轻野火相关损害的一个关键方面是在火灾发展到灾难性程度之前及早发现火灾。目前的做法是基于昂贵的资产，如卫星、了望塔和遥控飞机，这些资产需要不断的人力监督，将其使用限制在高风险或高价值地区。该项目旨在利用小型无人机（UAV）中计算，存储，传感和通信的超收敛性，实现对温度，植被，压力和化学浓度等环境因素的大规模测绘，这些因素有助于引发火灾。无人机团队可以自主充电，并在彼此之间以及与静态传感器进行间歇性通信，这是一个引人注目的研究目标，将有助于消防员进行持续的实时监视和早期发现随后发生的火灾。该提案提供了三项基本创新，以解决与自主协作环境监测相关的科学挑战。首先，提出了一种新的可满足性模最优控制框架，以处理混合的连续飞行动力学和离散约束，并确保避免碰撞，持续通信，和自主充电的无人机导航。其次，将开发一种使用新的不确定性加权模型的分布式系统架构，以实现跨无人机和静态传感器的异构团队的协作映射，并避免带宽密集型数据流。最后，提出了一种新的贝叶斯学习和推理方法来生成多模态（例如，具有自适应精度和不确定性量化的实时环境条件的热、语义、几何、化学）地图。该项目专注于多机器人团队的好处，例如，保护管理和搜救行动。这两个应用都要求机器人协调、合作和自主，包括多模态映射、异构网络上的协作推理以及具有安全、通信和能源约束的多目标导航。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Wildfires destroy millions of hectares of forest, sensitive ecological systems, and human infrastructure. A critical aspect of mitigating wildfire-related damages is early fire detection, well before initiating fires grow to disastrous proportions. Current practices are based on expensive assets, such as satellites, watchtowers, and remote-piloted aircraft, that require constant human supervision, limiting their use to high-risk or high-value areas. This project aims to take advantage of the hyperconvergence of computation, storage, sensing, and communication in small unmanned aerial vehicles (UAVs) to realize large-scale mapping of environmental factors such as temperature, vegetation, pressure, and chemical concentration that contribute to fire initiation. UAV teams that recharge autonomously and communicate intermittently among each other and with static sensors is a compelling research objective that will aid firefighters with continuous real-time surveillance and early detection of ensuing fires.This proposal offers three fundamental innovations to address the scientific challenges associated with autonomous, collaborative environmental monitoring. First, a new Satisfiability Modulo Optimal Control framework is proposed to handle mixed continuous flight dynamics and discrete constraints and ensure collision avoidance, persistent communication, and autonomous recharging for UAV navigation. Second, a distributed systems architecture using new uncertainty-weighted models will be developed to enable cooperative mapping across a heterogeneous team of UAVs and static sensors and avoid bandwidth-intensive data streaming. Lastly, a new Bayesian learning and inference approach is proposed to generate multi-modal (e.g., thermal, semantic, geometric, chemical) maps of real-time environmental conditions with adaptive accuracy and uncertainty quantification. This project with its focus on multi-robot teams benefits, e.g., conservation management and search-and-rescue operations. Both applications demand robot coordination, cooperation, and autonomy, including multi-modal mapping, collaborative inference over heterogeneous networks, and multi-objective navigation with safety, communication, and energy constraints.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.