3-Second Prediction of Urban Wind Direction! A team of Chinese scholars has developed a new AI technology to make drone flights more energy-efficient.
If a delivery drone needs to deliver a package several streets away, flying as straight as possible while avoiding buildings might seem like the most economical choice. But what if that straight path happens to run directly into a strong crosswind, crossing a particularly wide street? Should the drone fly straight through, veer slightly to the side, or change altitude? Is the shortest route necessarily the most energy-efficient? Recently, researchers from institutions including the University of Tennessee, Concordia University, McGill University, and Caltech published a paper introducing the GeoWind2Plan framework, which integrates urban three-dimensional wind field prediction with energy-efficient drone route planning. This work was carried out by Shaoxiang Qin (a visiting scholar in Professor Xiao’s group), Yucheng Zhao (a PhD student in Professor Xiao’s group), and others, with Professors Xiongye Xiao of the University of Tennessee, Liangzhu Wang of Concordia University, and Anima Anandkumar of Caltech serving as co-corresponding authors. (Source: Xiongye Xiao (DUAL)...)
