A Learning from Demonstration (LfD) Framework for Enhancing MASCOT Autonomy to Support Operators' Needs

增强 MASCOT 自主性以支持运营商需求的示范学习 (LfD) 框架

• 批准号: 2758997
• 金额: --
• 依托单位: Swansea University
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2758997
• 关键词: Learning; Demonstration; LfD; Framework; Enhancing

The project aims: 1. To conduct experiments where human operators manually control MASCOT robotic arm inside the fusion reactor and collect the usage data, 2. To use this data and apply learning from demonstration (LFD) methods that enable the MASCOT robot arm to move inside the fusion reactor in an autonomous or semi-autonomous manner.3. To conduct experiments to understand whether the LfD minimise human operators' workload and improve task efficiency. 4. Lastly, to investigate how operators' trust establishes over time with the integration of a semi-autonomous MASCOT robotic arm.

该项目的目标：1。为了进行实验，人类操作员手动控制聚变反应堆内的MASCOT机械臂并收集使用数据，2。使用这些数据并应用从演示中学习（LFD）的方法，使MASCOT机器人臂能够以自主或半自主的方式在聚变反应堆内移动。进行实验以了解LfD是否最大限度地减少了人类操作员的工作量并提高了任务效率。4.最后，研究如何随着时间的推移建立与半自主MASCOT机器人手臂的集成操作员的信任。