AI赋能城市轨道交通多专业综合和智能维保

陆雪忠

AI Empowering Multi-disciplinary Integrated and Intelligent Maintenance in Urban Rail Transit

LU Xuezhong
  • 作者信息:
    上海电气自动化集团有限公司副总裁, 正高级工程师
  • Affiliation:
    Vice President of Shanghai Electric Group Co., Ltd., Professor-level Senior Engineer
  • 栏目/Col:
    时评
时评 / Commentary
当今时代,AI浪潮扑面而来,千行百业席卷其中,纷纷寻找变革机遇、提升管理效率。我国城市轨道交通运营里程全球第一,规模庞大、场景众多,应用推广AI技术有着天然的基础和优势。在运营安全性、设备可靠性和维保经济性要求日益提高的背景下,城市轨道交通行业正积极探索和利用AI技术赋能,提高设备设施维保能力和水平,有效降低运维成本。当前,AI等新型信息化技术进一步推动从数字化到智能化的转型发展,也给城市轨道交通行业带来技术迭代、产业变革的重大机遇,成为高质量、可持续发展的核心引擎。“AI+城轨”全场景深度融合、AI赋能“智慧城轨”正式迈入2.0时代。车站机电设备包括站台门、电扶梯和自动售检票等众多专业,直接关系到乘客出行安全和服务体验,越来越受到关注和重视。同时,车站机电设备维保普遍存在“六多一高”现象(专业散而多、设备数量多、故障总量多、委外单位多、维保人员多、乘客投诉多、总体成本高),长期以来已成为行业共同的痛点和难点。上海电气自主研发的“车站机电多专业综合&智能维保(iPHMart-S)系统和整体解决方案”,具有自主化、智能化、绿色化,以及小而精、轻量化、部署快捷等特点和优势。与传统的单专业维保模式相比,设备管理水平明显提升,综合成本下降30%以上。1) 一个数字底座。包括设备数字台账和规程结构化、数字化台账。按设备专业系统重要度分为A、B、C三个等级,制订不同等级的台账颗粒度标准和维修策略,包括状态修、计划修和故障修。针对不同规程,按不同维护频次将各专业进行结构化和数字化,并和设备数字台账形成一一映射关系。2) 两个功能模块。包括计划管理模块和故障管理模块。计划管理模块指按计划实施的巡检维护作业,以计划工单形式,由巡检维护班组实施,并由车站运营部门确认形成闭环;故障管理模块指从不同来源的故障报修,以故障工单形式,由维修抢修班组负责实施,形成故障闭环。3) 一套设备履历。基于计划工单和故障工单的实施,实现全过程管理闭环,形成完整的设备电子履历。随着时间推移和数据积累,既可有效避免纸质工单分散保存、易遗失、难追溯等共性难题,又可打造真正意义上的设备运维数据资产基础,为多种应用提供数据支撑。4) 标准化组织架构。采用扁平化组织架构(1+3)和检修分离模式,设置1个中心(MCC——数字化运维协调中心)+3个班组(技术支持和数据分析班组、维修抢修班组、巡检维护班组)。人员采用标准化配置:巡检维护班组按规程工时配置,维修抢修班组按车站数配置,技术支持班组按专业数配置,MCC中心按轮值班次配置。从而实现人员配置定额化、工时管理人性化、成本控制透明化、维保作业高效化。5) 一套AI分析体系。基于iPHMart-S系统的技术支撑,在自动建立设备故障履历的同时,每天将产生大量标准格式的设备状态、故障和人员工时数据,为AI实时分析提供基础条件。通过技术支持班组的数据统计和专业分析,便可轻易获得:哪些车站、哪些系统、哪些时段、哪些备件、哪些人员发生了哪些情况,需要什么样的资源支撑,从而可实现人员精准化配置和培训、设备精细化管理和提升、成本透明化管控和压降、模式标准化推广和应用。截至2024年底,我国城市轨道交通运营里程超12 000 km,车站数超6 650座。按平均每站每年维保费用100万元测算,年市场规模达66.5亿元。按综合成本节约30%估算,合计可节约20亿元/年。一升一降,“小系统”破解行业大难题。AI赋能城市轨道交通高质量发展,已经从概念探索进入实际落地阶段。上海电气将依托在城市轨道交通装备领域的产业基础和数字化、智能化优势,为广大客户提供行业领先的高效率、低成本的“AI+城轨”整体解决方案。
The wave of AI (artificial intelligence) is surging across industries in this era, with all sectors seeking opportunities for transformation and improvement in management efficiency. The urban rail transit network in China ranks first globally in terms of operating mileage, with vast scale and numerous application scenarios, providing a natural foundation and advantage for the application and promotion of AI technologies. In the context of increasing requirements for operational safety, equipment reliability, and maintenance cost efficiency, the urban rail transit industry is actively exploring and leveraging AI to enhance the capabilities and standards of equipment maintenance, thereby effectively reducing operation and maintenance costs.At present, AI and other emerging information technologies are further driving the transformation from digitalization to intelligence, bringing major opportunities for technological iteration and industrial transformation in the urban rail transit sector. AI has become a core engine for high-quality and sustainable development. The deep integration of ′AI+Urban Rail′ across all scenarios marks the official entry into the 2.0 era of AI-empowered ′Smart Urban Rail.′Station electromechanical systems include numerous disciplines such as platform screen doors, escalators, and automatic fare collection, which are directly related to passenger travel safety and service experience, receiving increasing attention. However, station electromechanical maintenance commonly faces the ′six-many and one-high′ phenomenon (referring to many sporadic disciplines, many pieces of equipment, many failures, many outsourced contractors, many maintenance staff, many passenger complaints, and high overall costs). This has long been a common challenge and difficulty across the industry.Shanghai Electric has independently developed the ′Station Electromechanical Multi-disciplinary Integrated & Intelligent Maintenance (iPHMart-S) System and Comprehensive Solution,′ characterized by autonomy, intelligence, eco-friendliness, compact and refined design, lightweight structure, and rapid deployment. Compared with traditional single-specialty maintenance mode, this solution significantly improves equipment management levels while reducing overall costs by more than 30%.1) One digital foundation. This includes digital equipment ledgers and structured maintenance procedures, and digitized ledgers. According to the importance of professional equipment systems, they are classified into three levels—A, B, and C. Different ledger granularity standards and maintenance strategies are formulated for each level, covering condition-based maintenance, planned maintenance, and corrective maintenance. For various procedures, disciplines are structured and digitized according to maintenance frequencies, forming a one-to-one mapping relationship with the digital equipment ledger.2) Two functional modules. These include the planning management module and the fault management module. The planning management module refers to inspection and maintenance tasks carried out according to plan, executed in the form of planned work orders by inspection and maintenance teams, and confirmed by station operation departments to form a closed-loop process. The fault management module refers to fault repair requests originating from different sources, executed in the form of fault work orders by emergency repair teams, thereby forming a closed loop for fault scenarios.3) One set of equipment records. Based on the execution of planned and fault work orders, a closed-loop process of full-cycle management is achieved, forming a complete electronic equipment record. With the passage of time and accumulation of data, this not only effectively avoids common problems such as scattered storage, loss, and difficulty in tracing paper-based work orders, but also establishes a true foundation of equipment operation and maintenance data assets, providing strong data support for a wide range of applications.4) Standardized organizational structure. A flat organizational structure (1+3) and a maintenance-inspection-separated model are adopted, consisting of one center (MCC—digital operation and maintenance coordination center) plus three teams (technical support & data analysis team, emergency repair team, and inspection & maintenance team). Personnel are allocated in a standardized manner: the inspection & maintenance team is configured based on procedure man-hours, the emergency repair team is configured according to the number of stations, the technical support team is configured by specialty, and MCC is configured by duty shifts. This enables standardized personnel allocation, humanized man-hour management, transparent cost control, and efficient maintenance operations.5) One AI analysis system. Supported by the iPHMart-S system, while automatically generating equipment fault records, large volumes of standardized data on equipment status, failures, and personnel man-hours are produced daily, providing the foundation for real-time AI analysis. Through statistical work and professional analysis conducted by the technical support team, it becomes easy to identify: which stations, which systems, which time periods, which spare parts, and which personnel are associated with what specific situations, and what resource support is required. This enables precise personnel allocation and training, refined equipment management and enhancement, transparent cost control and reduction, as well as standardized model promotion and application.By the end of 2024, China′s urban rail transit operating mileage exceeded 12 000 km, with more than 6 650 stations. Based on an average annual maintenance cost of RMB 1 million per station, the annual market size reaches RMB 6.65 billion. With an estimated 30% reduction in comprehensive costs, annual savings could amount to RMB 2 billion.One rise, one fall: a ′small system′ addresses major industry challenges. AI empowerment is propelling the high-quality development of urban rail transit, transitioning from conceptual exploration to practical implementation. Leveraging its industrial foundation in urban rail transit equipment, along with its strengths in digitalization and intelligence, Shanghai Electric will provide customers with industry-leading, high-efficiency, low-cost ′AI+Urban Rail′ integrated solutions.
Translated by ZHANG Liman
论文检索