Automated Model-Based Sensor Planning

9311877 Allen这是该奖项连续三年颁发的第一年。研究的重点是自动确定传感器参数值的策略，以规划传感器任务。这将使价值选择和定制能够匹配任务要求，从而支持传感器系统更灵活、更可靠的自主操作。在哥伦比亚大学已经实施的机器视觉规划系统（MVP）的基础上，新的研究将扩展这一框架，并支持在静态和动态环境中使用该系统进行实验。研究的三个重点是：1)提高优化方法的性能，以寻找可接受的观点；2)将静态传感器规划结果扩展到动态环境；3)将MVP扩展到视觉以外的传感器系统（触觉、距离和接近）。

9311877 Allen This is the first year of a three-year continuing award. The research focuses on strategies for automatically determining sensor parameter values for planning of sensor tasks. This will enable value selection and tailoring to match task requirements, thereby supporting more flexible and reliable autonomous operation of sensor systems. Building on a Machine Vision Planning system (MVP) already implemented at Columbia University, the new research will extend this framework, and support experiments with the system in static and dynamic environments. Three main thrusts of the research are: 1) improving the performance of the optimization approach to finding admissible viewpoints; 2) extending static sensor planning results to dynamic environments;and 3) extending MVP to sensor systems other than vision (tactile, range, and proximity).