NCS-FO: Advantages of varying navigational abilities in humans and robots

尽管导航技能很重要，但人们找到路的能力和采取的方法却千差万别。然而，我们不知道为什么会存在这种差异，也不知道在日常导航中如何使用不同的策略。在机器人、自动驾驶汽车和其他自动系统中，越来越明显的是，一刀切的方法并不适用于所有环境和用户需求。同样，生产具有不同导航强度的自主系统可以提高自主系统的整体能力，例如探索或搜索和救援机器人团队。该项目汇集了来自神经科学，认知心理学，计算机科学和机器人技术的研究人员，以研究复杂环境中导航能力和策略的变化。研究人员结合了联合收割机行为，神经科学和计算方法。精确定位个体差异背后的神经和行为标记将为导航挑战提供定制的解决方案，并优化自主系统在不同环境条件下的性能。这项研究的成果将对社会和科学发现产生以下具体效益：1）推进对空间导航过程的理论理解，2）理解人类和机器人的差异优势，影响人们如何接近这两个领域，3）对自动驾驶汽车，GPS寻路设备和交通标志的改进的影响，本研究的主要跨学科目的是：1）检验人类空间导航能力是单一能力还是多种能力的贡献，2）通过多模态成像建立人类导航能力的神经标记，以及3）在真实世界的情况下实现和测试机器人的不同导航能力。这项研究将是迄今为止最大的（n = 270）研究人类导航能力，使用结构方程模型和成像数据的多变量分析来深入解决这个问题。此外，该项目将扩大能力的范围，将导航技能与工作记忆，学习，个性和其他因素联系起来。在机器人中实现导航策略和能力提供了对它们权衡的受控测试。通过操纵机器人的规划和映射策略，研究人员可以分离出特定的能力及其对导航的贡献，测试自主系统的导航理论和实际优势。该项目的跨学科方法利用认知科学，机器人和神经科学的优势来测试人类个体差异的基本性质。该奖项反映了NSF的法定使命，并被认为值得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来支持。

Despite the importance of navigational skill, people’s ability to find their way around and the approaches they take vary widely. Yet we do not know why this variation exists or how different strategies are used during everyday navigation. In robotics, self-driving cars, and other autonomous systems, it has become increasingly clear that a one-size-fits-all approach is not viable for all environments and user needs. Similarly, producing autonomous systems with different navigational strengths could improve the capacity of autonomous systems as a whole, such as teams of exploring or search-and-rescue robots. This project brings together researchers from neuroscience, cognitive psychology, computer science, and robotics to study variability in navigation abilities and strategies in complex environments. The researchers combine behavioral, neuroscience, and computational approaches. Pinpointing the neural and behavioral markers that underlie individual differences will lead to customized solutions to navigational challenges and optimize the performance of autonomous systems for differing environmental conditions. The outcomes of this research will have the following specific benefits to society and scientific discovery: 1) advancing theoretical understanding of the processes involved in spatial navigation, 2) understanding the advantages of variation in both humans and robots, impacting how people approach both fields, 3) implications for improvements in self-driving cars, GPS wayfinding devices, and transportation signage, and 4) broader dissemination of virtual reality (VR) technology.The overarching cross-disciplinary aims of this study are to 1) test whether human spatial navigation is a singular competence or whether multiple abilities contribute, 2) establish the neural markers of human navigational abilities through multi-modal imaging, and 3) implement and test different navigational abilities in robots in real-world situations. This study will be the largest to date (n = 270) to study human navigation abilities, using both structural equation modeling and multivariate analysis of imaging data to deeply address this question. Furthermore, the project will broaden the scope of abilities to relate navigation skills to working memory, learning, personality, and other factors. Implementing navigational strategies and abilities in robots provides a controlled test of their tradeoffs. By manipulating the planning and mapping strategies of robots, the researchers can isolate particular abilities and their contributions to navigation, testing both navigational theory and practical advantages for autonomous systems. The interdisciplinary approach of this project harnesses the strengths of cognitive science, robotics, and neuroscience to test the fundamental nature of human individual variability.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.