Robotics and Artificial Intelligence for Nuclear Plus (RAIN+)

核+机器人和人工智能（RAIN）

• 批准号: EP/W001128/1
• 负责人: Barry Lennox
• 金额: $ 251.71万
• 依托单位: University of Manchester
• 依托单位国家: 英国
• 项目类别: Research Grant
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=EP%2FW001128%2F1
• 关键词: Robotics; Artificial; Intelligence; Nuclear; Plus

The nuclear industry has a vast array of highly complex and diverse challenges that span decommissioning, waste management, fission power plants, advanced modular reactors and fusion reactors. In the UK, one of the most significant challenges is to decommission legacy storage facilities. There is estimated to be approximately 3,000 tonnes of high-level waste (HLW), 310,000 tonnes of intermediate level waste (ILW) and hundreds of radioactive facilities that need to be decommissioned in the UK alone. Despite significant progress during the first phase of RAIN, decommissioning continues to rely almost exclusively on manual operations, requiring people to enter extremely hazardous environments placing themselves at risk. Significant amounts of personal protective equipment (PPE) is required, which reduces dexterity and lowers productivity to levels significantly below that of other industries. PPE also adds significantly to the waste materials that must be disposed of and as a consequence, makes some future operations infeasible. For example, it has been estimated that more than 1 million suited entries will be required to decommission the THORP plant alone on the Sellafield site. RAI technologies are therefore considered essential if the UK is to address its decommissioning challenges. In the future generation of nuclear power, fusion reactors will not be able to operate without advances being made to remote handling equipment. In addition, remote inspection and maintenance of new fission reactors is essential if they are to be commercially viable.RAIN+ will continue to push the boundaries of Robotics and AI (RAI) science, developing robotic solutions that solve major challenges facing the nuclear sector. To ensure that the work is relevant, has a long-term impact on industry, and leads to deployments of RAI technology into active facilities, RAIN will continue to work in close partnership with nuclear end-users, the supply chain and regulators. Furthermore, recognising that many of the hazards encountered in the nuclear industry are prevalent in other industry sectors, such as agriculture, construction, offshore and healthcare, RAIN will work to expand its user and application base such that RAI solutions can be developed that have cross-sector relevance and a single hub for all challenging environments, not just nuclear, can be established towards the end of this second phase.

核工业面临着一系列高度复杂和多样化的挑战，包括退役、废物管理、裂变发电厂、先进模块化反应堆和聚变反应堆。在英国，最重大的挑战之一是退役遗留存储设施。据估计，仅在英国就有大约3,000吨高放射性废物（HLW），31万吨中放射性废物（ILW）和数百个放射性设施需要退役。尽管在RAIN的第一阶段取得了重大进展，但退役仍然几乎完全依靠人工操作，需要人们进入极度危险的环境，使自己处于危险之中。需要大量的个人防护设备（PPE），这降低了灵活性，并将生产力降低到远低于其他行业的水平。个人防护设备还大大增加了必须处理的废料，因此，使一些未来的操作变得不可行。例如，据估计，仅在塞拉菲尔德现场的THORP工厂就需要100多万件套装。因此，如果英国要解决其退役挑战，RAI技术被认为是必不可少的。在未来的核能发电中，如果没有远程处理设备的进步，聚变反应堆将无法运行。此外，远程检查和维护新的裂变反应堆是必不可少的，如果他们是商业可行的。RAIN+将继续推动机器人和人工智能（RAI）科学的边界，开发机器人解决方案，解决核领域面临的主要挑战。为了确保这项工作具有相关性，对行业产生长期影响，并将RAI技术部署到现役设施中，RAIN将继续与核最终用户，供应链和监管机构密切合作。此外，认识到核工业中遇到的许多危险在其他工业部门也很普遍，如农业，建筑，海上和医疗保健，RAIN将努力扩大其用户和应用基础，以便开发具有跨部门相关性的RAI解决方案，并在第二阶段结束时为所有具有挑战性的环境建立一个单一的中心，而不仅仅是核能。

项目成果

CART-I: Design and Development of Collision Avoidance Robotic Tether for Underwater Nuclear Inspection Vehicles

CART-I：水下核检查车防撞机器人系绳的设计与开发

• DOI: 10.23919/oceans44145.2021.9705689
• 发表时间: 2021
• 作者: Al Mhdawi A

Fault-tolerant Control of Robotic Systems with Sensory Faults using Decoupled Active Inference

使用解耦主动推理对具有感知故障的机器人系统进行容错控制

• 发表时间: 2021
• 作者: Baioumy M

Task Learning for Intention Detection using Deep Neural Networks and Robotic Arm Data in Glovebox

使用深度神经网络和手套箱中的机械臂数据进行意图检测的任务学习

• 发表时间: 2022
• 作者: Alharthi A

Risk-Aware Motion Planning in Partially Known Environments

部分已知环境中的风险感知运动规划

• 发表时间: 2021
• 作者: Barbosa F

Autonomous Systems' Safety Cases for use in UK Nuclear Environments

用于英国核环境的自主系统安全案例

• DOI: 10.4204/eptcs.391.10
• 发表时间: 2023
• 期刊: Electronic Proceedings in Theoretical Computer Science
• 作者: Anderson C