The Ministry of Transport issues the Guidelines for Artificial Intelligence Application Scenarios in the Water Transport Sector.

Title context: The Ministry of Transport issues the Guidelines for Artificial Intelligence Application Scenarios in the Water Transport Sector. Text: Recently, the General Office of the Ministry of Transport issued the Guidelines for Artificial Intelligence Application Scenarios in the Water Transport Sector, proposing 22 AI application scenarios in the water transport sector with urgent needs from three aspects: safety enhancement, quality and efficiency improvement, and green transformation. In terms of key scenarios for quality and efficiency improvement, 10 typical application scenarios were proposed in areas such as automated terminal operations, smart collaboration in port logistics, smart waterway operation services, and digital and intelligent shipping operation services, to promote the accelerated deep integration of digital and intelligent technologies with application scenarios in relevant business areas and enhance the efficiency and service quality of waterway transport. In terms of key scenarios for green transformation, 3 typical application scenarios were proposed in areas such as green and intelligent port operations, green and intelligent shipping management and control, and green and compliant intelligent supervision of ships, to promote the deepened integrated application of green technologies and artificial intelligence in relevant business areas and empower the green and sustainable development of the industry. In terms of key scenarios for safety enhancement, 9 typical application scenarios were proposed in areas such as intelligent prevention and control of port safety risks, intelligent assurance of safe waterway operations, intelligent prevention and control of shipping operation risks, intelligent monitoring of water traffic safety, and enhancement of intelligent maritime search and rescue capabilities, to promote the strengthening of precise risk prediction and dynamic prevention and control in relevant business areas and enhance safety assurance and emergency response capabilities. The original text is as follows: Guidelines for Artificial Intelligence Application Scenarios in the Water Transport Sector I. Key Scenarios for Safety Enhancement (1) Intelligent safety management for port enterprises. Build a safety management agent, relying on knowledge graphs, large language models, and process automation technologies to achieve intelligent verification of safety compliance, independent troubleshooting of hidden dangers, and online personalized education and training, and improve the level of intelligent management and control of port safety. Focusing on typical high-risk scenarios such as human-machine cross-operation, dangerous goods operations at the ship-shore interface, hot work and confined space operations, and operations in severe weather, build scenario-based risk deduction and quantitative assessment models, integrate video recognition, multi-source data collection, and intelligent perception technologies, and achieve early identification, dynamic monitoring, intelligent analysis, and intelligent early warning of safety risks. (2) Monitoring and early warning of safety risks for dangerous goods storage tanks at ports. Focusing on major hazard source storage tanks for dangerous goods at ports, integrate multi-source monitoring data such as temperature, pressure, liquid level, combustible and toxic gas alarms, and intelligent video recognition, build a safety risk monitoring and early warning agent, conduct real-time perception, intelligent analysis, and proactive early warning for typical risks such as abnormal operating parameters of storage tanks, leakage of combustible and toxic media, and violations in personnel operations, and improve the intrinsic safety level of major hazard source storage tanks at ports. (3) Monitoring and early warning for the healthy operation and safety of large port loading and unloading equipment. Use intelligent inspection Siasun Robot&Automation, sensor networks, and other means to collect multi-source data such as vibration, temperature, and voiceprints, build intelligent fault diagnosis and prediction models, achieve real-time perception, intelligent analysis, trend prediction, precise early warning, and remaining useful life assessment of equipment operating status, promote the transformation of operation and maintenance models from reactive maintenance after failures to preventive and predictive maintenance, and comprehensively improve the intrinsic safety level of large port loading and unloading equipment. (4) Intelligent and safe operation and maintenance of inland waterway navigation structures. Use drones, underwater Siasun Robot&Automation, and intelligent sensor networks, integrate multi-source monitoring data, develop an intelligent agent for operation and maintenance inspection and early warning of navigation structures, achieve precise identification and autonomous assessment of the health status of navigation structures, analyze abnormal change trends of facilities, support preventive maintenance of navigation structures such as hydraulic structures, gates and valves, and hoists, enhance the effectiveness of safety supervision in lock hub areas, and improve navigation support capacity. (5) Intelligent deduction and prediction of underwater topography in inland waterways. Focusing on riverbed evolution and prediction in inland waterways under the operating conditions of large hub groups, integrate data such as waterway water levels and historical riverbed evolution, develop along-route water level prediction methods and underwater topography deduction and prediction models that integrate physical mechanisms and machine learning, achieve timely mastery and analytical deduction of key waterway status data such as navigation trajectories, ship positions, cross-sectional water depths, siltation volumes and locations, and target characteristics of underwater navigational obstacles during survey processes, improve the ability to predict waterway topographic evolution, and provide support for waterway dredging planning and navigation safety assessment. (6) Ship navigation safety monitoring and emergency response. Integrate monitoring data from waterborne video surveillance systems, automatic identification systems, vessel traffic management systems, and others, develop a multimodal vertical-domain model for shipping safety, and achieve ship navigation safety monitoring, intelligent identification of abnormal ship behavior, and traffic situation analysis and early warning in complex waterway sections. Build a shipping safety emergency knowledge base and knowledge graph, combine vertical-domain models to construct a shipping safety emergency agent, achieve automatic generation of accident response emergency plans and intelligent allocation of emergency resources, and improve waterborne safety emergency command capabilities. (7) Intelligent analysis of international shipping market risks. Build risk identification models covering key shipping corridors and important hub ports, develop a "corridor-port-ship" comprehensive analysis agent, and conduct dynamic perception and intelligent analysis of abnormal ship navigation, anchoring, and operation behavior. Gather multi-source data such as international shipping rules and market information, carry out intelligent information analysis and assessment, support information tracking, compliance management, and policy evaluation, and improve the foresight and compliance of enterprise operations. (8) Network and data security assessment of shipborne intelligent systems. In response to new risks after the deep embedding of artificial intelligence technologies in scenarios such as navigation decision-making, predictive maintenance, and ship-shore collaboration, use security testing technologies such as adversarial samples and data poisoning to quantitatively assess the stability of models such as intelligent collision avoidance and fault diagnosis, analyze risks such as malicious manipulation, data theft, and output distortion in different scenarios, enhance ship resilience against attacks, and improve navigation safety. (9) Intelligent search and rescue for people in distress at sea. Deepen key technology research on unmanned intelligent rescue and salvage equipment, promote the deployment and application of new intelligent equipment such as satellite remote sensing, drones, unmanned surface vessels, and unmanned underwater vehicles in waterborne and underwater search and rescue operations, continue to advance the improvement and upgrading of algorithms for small and weak targets at sea, and improve the efficiency of search and rescue for people in distress at sea. II. Key Scenarios for Quality and Efficiency Improvement (10) Unmanned operations at automated terminals. Comprehensively apply technologies such as BeiDou satellite navigation, 5G communications, and artificial intelligence, establish an integrated collaborative management and control platform for the full process of port machinery and equipment, and promote the unmanned development of ship berthing and unberthing, quayside loading and unloading, horizontal transport, and yard operations. (11) Smart external services at ports. Build high-quality port logistics datasets and knowledge graphs, rely on technologies such as parameter fine-tuning and knowledge enhancement to create a vertical-domain large model for logistics. Develop intelligent agents for typical scenarios such as port public services, intelligent customer service for logistics chains, and end-to-end logistics route decision-making, build an integrated port logistics cloud service platform, achieve intelligent scheduling and full-process status visibility across the entire port collection and distribution chain, improve the intelligence level of port external services, and promote cost reduction and efficiency improvement in logistics. (12) Intelligent organization and scheduling of rail-water intermodal transport. Build a dynamic adaptive model for integrated collaborative scheduling of "railway-port-shipping," focus on core transport links or nodes such as railway freight cars and stations, port terminals and berths and warehouses, inbound and outbound port channels, and ships, and achieve dynamic monitoring, coordinated control, and global optimization of full-process operation plans and operating status. Deepen the application of "single document" and "single container" systems, and improve the organizational efficiency of rail-water intermodal transport and the smoothness of the supply chain. (13) Intelligent scheduling of multi-level ship locks. Integrate multi-source data such as ship lockage demand, ship lock equipment conditions, navigation environment conditions, and lock chamber dimension parameters, build an intelligent agent for multi-level ship lock operation scheduling, rely on edge computing nodes and BeiDou shipborne terminals, achieve ship lockage demand prediction, intelligent arrangement and scheduling, dynamic decision optimization for joint scheduling of multi-stage ship locks, and efficient utilization of lock chamber space, and comprehensively improve waterway navigation support capacity and the level of safe operation and control of ship locks. (14) Accurate calculation of crowdsourced water depths in inland waterways. Focusing on accurate prediction of waterway water depth, couple ship basic data, monitoring data from shipborne depth sounding equipment, ship navigation attitude data, and waterway navigation environment data, build a correction model for ship crowdsourced depth sounding data, achieve automatic spatiotemporal matching of ship trajectories and depth sounding data, intelligent correction and integrated application of ship crowdsourced depth sounding data, effectively improve the utilization efficiency of ship crowdsourced water depth data, and reduce waterway surveying and mapping costs. (15) Intelligent operation of inland river online freight platforms. Promote inland river freight transport online platform operators to use technologies such as big data, blockchain, and artificial intelligence to accelerate platform intelligent upgrading, optimize functions such as ship-cargo matching, payment settlement, and risk control early warning, promote the full-process online and transparent nature of inland river transport services, and improve the organizational efficiency and transaction security of inland river online freight transport. (16) Intelligent container allocation for shipping enterprises. Integrate multi-source data such as routes, capacity, port dynamics, and historical container usage, build an empty container scheduling agent based on reinforcement learning and operations optimization, carry out precise forecasting of global container usage demand, automatically generate empty container repositioning plans under constraints such as container usage assurance and timeliness, achieve end-to-end intelligent allocation across overseas main lines, domestic trade routes, and overseas regional areas, and improve the utilization efficiency of global container resources. (17) Intelligent analysis of economic activities of shipping enterprises. Build a dynamic monitoring model for the compliance of economic activities of shipping enterprises, and achieve dynamic monitoring of compliance status and advance prevention and control of compliance risks according to external regulatory requirements and internal operation and management needs. Build an intelligent agent for monitoring the level of shipping variable costs, conduct real-time collection and aggregation of expenses such as voyage-related, operational, business-related, and compliance-related costs, generate refined continuous improvement plans for operations through budget benchmarking, historical benchmarking, and industry benchmark benchmarking, and achieve abnormal early warning and cost optimization. (18) Intelligent optimization of international liner routes. Focusing on international liner route network planning and operations, integrate multi-source data such as freight demand and port call conditions, build an intelligent route network planning model, conduct intelligent analysis, scientific deduction, and dynamic optimization of key elements such as route design, port call sequence, ship type configuration, schedule arrangement, and cargo flow paths, and improve the operating efficiency and service reliability of international liner route networks. (19) Trusted blockchain applications in shipping logistics. Promote the expansion and increase of digital shipping trade, accelerate the trusted circulation, aggregation, and intelligent application of logistics data, deepen the application of business scenarios such as electronic bills of lading and electronic delivery orders, build an industry electronic document chain based on intelligent document recognition technology, promote cross-platform mutual recognition and alignment with international standards, and further enhance collaborative service capabilities in port and shipping business. Promote the connection of electronic bill of lading systems with the DIGIHUMAN RMB blockchain service platform, and enhance the capacity for domestic settlement of ocean freight and DIGIHUMAN RMB settlement. III. Key Scenarios for Green Transformation (20) Green and low-carbon intelligent management and control at ports. Build a green and low-carbon intelligent management and control platform for ports, deploy energy consumption monitoring sensor networks, carry out real-time monitoring and intelligent analysis of factors such as energy consumption, carbon emissions, and clean energy, optimize port area energy scheduling and energy use structure, promote the integrated application of clean energy such as photovoltaics, energy storage, and shore power, improve the utilization efficiency of renewable energy, and achieve calculable port carbon emissions and controllable energy consumption. (21) Green and intelligent planning of international shipping capacity. Develop a question-answering agent covering domestic and foreign carbon emission policies, achieve intelligent interpretation of policy provisions, and consolidate the foundation for enterprise compliance management. Build an intelligent calculation model for international shipping capacity allocation and cost-effectiveness adapted to policy requirements, use integrated analysis of multi-source shipping big data, deep reinforcement learning, and multi-objective intelligent optimization algorithms, achieve intelligent comparative calculation of ship capacity allocation under carbon policy constraints and cost-effectiveness under different operating modes, and improve the level of green and intelligent planning of enterprise shipping capacity. (22) Intelligent supervision and services for shipping carbon emissions. Coordinate ship energy consumption, enterprise operating data, and carbon management and control policies, build intelligent assessment models for dynamic accounting of shipping carbon emissions and ship energy efficiency indicators, intelligently generate carbon emission accounting reports, performance cost calculation and assessment, and differentiated emission reduction pathway plans for shipping enterprises, provide intelligent support for carbon emission data verification and carbon intensity rating for maritime management agencies, and improve carbon compliance management and control capabilities. This article is edited from the official website of the "Ministry of Transport." GMTEight editor: Jiang Yuanhua.