I-Corps: Real-time intelligent sensor path planning based on information value estimation

I-Corps项目更广泛的影响和商业潜力在于开发可靠的自动驾驶汽车系统，该系统可以通过利用传感器测量来适应车辆在行驶过程中的路径，从而克服现有的限制。由于雷达、激光雷达和声纳传感器可以在机载机器人平台上使用，市场可能会利用自主的空中、水下和地面车辆。其中一个应用是开发更可靠、更安全的自动驾驶系统。另一个例子是自主空中/水下航行器，它可以帮助收集野生动物栖息地和种群的高质量数据，同时最大限度地减少对自然环境的破坏。确定现有自主系统的要求将有助于为自主机器人系统的可靠设计建立一个通用标准。I-Corps的这个项目是基于传感器路径规划算法的开发，该算法可以在车辆行驶时根据传感器测量值调整其方向。该算法采用信息理论和概率传感器测量模型来确定车辆的路径，从而获得最具信息量的测量结果。该算法能够考虑基于目标识别的置信度，车辆可以根据实时收集到的信息自动重新规划路径。该技术的基础是研究有效和高效的方法来利用和处理传感器的测量结果进行路径规划。技术结果表明，该算法可以计算出最小时间路径，同时获得令人满意的传感器测量置信水平。路径规划算法的应用将赋予更高的自主性，从而形成更可靠的集成系统。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact/commercial potential of this I-Corps project is the development of reliable autonomous vehicle systems that may overcome existing limitations by utilizing sensor measurements to adapt its path while the vehicle is in motion. As RADAR, LiDAR, and SONAR sensor can be used on the onboard robotic platforms, markets may utilize autonomous aerial, underwater, and ground vehicles. One application is the development of a more reliable and safe autonomous driving system. Another example is an autonomous aerial/underwater vehicle that may assist in gathering quality data of wild animal habitats and populations while minimizing the damage of natural environment. Defining the requirements of the existing autonomous systems will help establish a general standard for a reliable design for autonomous robotic systems.This I-Corps project is based on the development of a sensor path planning algorithm that may adapt its direction based on sensor measurements while the vehicle is in motion. Information theory and probabilistic sensor measurement models are used in the algorithm to decide the vehicle’s path such that the most informative measurements can be obtained. The algorithm is capable of considering the confidence level based on object identification and the vehicle can automatically re-plan its path based on the information gathered in real-time. This technology is based on research on effective and efficient methods to utilize and process the sensor measurements towards path planning. Technical results demonstrate that the proposed algorithm can compute the minimum-time path while achieving a satisfactory confidence level from obtained sensor measurements. The application of the path planning algorithm will give a higher level of autonomy, resulting in a more reliable integrated system.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.