NetWork
Developmental logic，functional positioning， and paradigm innovation of philosophy and social science laboratories in the digital-intelligence Era
LI Deli;XIONG Yan;[Objective]Philosophy and Social Science Laboratories (PSSLs) play a critical role in advancing the construction of New Liberal Arts， integrating social science resources， and driving research paradigm innovation. However， there remains no unified consensus on their functional positioning， while discussions on developmental paradigm innovation lack systematic analysis. This study proposes a dual-path development logic for PSSLs: "Critical Inheritance-Innovation through Historical-Realistic Evolution" and "Techno-Social Synergistic Co-Creation". It further clarifies PSSLs' functional positioning in: (1) cultivating interdisciplinarily innovative talents， (2) enabling cross-disciplinary scientific research innovation， (3) facilitating government-industry-academia-research interaction and translation， (4) supporting policy advisory and decision-making， and (5) promoting the global dissemination of Chinese discourse. A paradigm innovation framework encompassing data spaces， organizational development， pedagogical practices， scientific research， and knowledge production is constructed to provide insights for high-quality development of PSSLs in the Digital-Intelligence Era.[Methods]This study employed a literature analysis methodology to comprehensively review the current domestic research landscape on PSSLs. Through critical evaluation of existing scholarship， key research gaps were identified， leading to the proposition of PSSLs’ five core functional positioning dimensions:Cultivation of interdisciplinary innovative talents，Cross-disciplinary scientific research innovation，Government-industry-academia-research mutual translation，Policy advisory support for decision-making，Global dissemination of Chinese discourse.Subsequently， case analysis methodology was applied to examine developmental paradigm innovation in PSSLs. The research demonstrates that high-quality development of PSSLs can be advanced through innovations across five paradigm domains:Data spatial paradigms，Organizational developmental paradigms，Educational pedagogical paradigms，Scientific research paradigms，Knowledge production paradigms.[Results]Research Findings: PSSLs constitute vital institutional carriers for accelerating the establishment of China's self-directed philosophy and social science knowledge system， serving as new platforms for disciplinary prosperity. Their digitization-driven and intelligence-empowered qualitative transformation reflects a dialectical unity of historical and realistic logic， emerging from the synergy between technological and societal logic. theoretically， PSSLs are capable of integrating resources across diverse academic fields， dissolving disciplinary barriers， and generating theoretical insights that align with China’s national conditions and social development patterns. This provides foundational support for constructing an autonomous Chinese knowledge system in philosophy and social sciences.practically， PSSLs address complex challenges arising in the process of national governance by employing scientific research methods and advanced technical tools. They deliver forward-looking， targeted， and actionable policy recommendations and solutions， thereby supplying intellectual support for modernizing China’s governance system and governance capacity.[Conclusions]PSSL construction transcends the mere adoption of advanced technological tools; it represents key infrastructure driving disciplinary convergence， methodological renewal， knowledge production transformation， and social value creation. This shift marks an epistemological transition from small-data hypothesis verification to big-data pattern discovery， and from empirical summarization to simulation-based scenario projection. The evolution of PSSLs embodies continuous methodological innovation and provides pragmatic solutions for addressing complex societal challenges and enhancing modern governance. Current challenges include technological ethics and disciplinary discourse reconstruction. Future development requires strengthening humanistic guidance over technological applications to achieve systemic responses to complex social problems.
Proteomics analysis of plasma with hepatocellular carcinoma based on label-free quantitative technology
LI Yilan;YU Wenjing;SHI Wenchao;LI Lan;[Objective] Early symptoms of hepatocellular carcinoma (HCC) are often occult， most patients are diagnosed at advanced stages with poor prognosis. Therefore， it is of great significance to deeply analyze the molecular mechanism of HCC occurrence and development and to search for sensitive early diagnostic biomarkers. Ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) plays a pivotal role in proteomic research. To solve the problem of low coverage in protein identification， which is caused by the significant differences in abundance and the complex composition of plasma proteins. The influence of different liquid chromatography gradient separation conditions on the identification of plasma proteins was investigated. Furthermore， the developed method was applied to the proteomic analysis of plasma samples from HCC patients and healthy individuals， with the aim of identifying early diagnostic biomarkers for cancer and providing support for the molecular mechanism of disease occurrence and development. [Methods] In this study， to investigated the effects of different liquid chromatography gradient separation conditions on the identification of plasma proteins， the liquid chromatography separation gradients were set at 56， 85， and 130 min. A standardized plasma proteomics analytical workflow suitable for use in a public research platform was ultimately established based on the results. Furthermore， combined with the label-free quantitative proteomics technology， differential expression proteins in plasma samples from HCC patients and healthy individuals were screened. Bioinformatics analyses， including molecular function and biological process， were performed to elucidate molecular signatures associated with HCC pathogenesis. [Results] After extending the separation gradient from 56 minutes to 85 minutes and 130 minutes， the number of identified plasma proteins increased from 588 to 988 and 1 261. Also， 64.1% and 63.6% of the proteins identified under the 56 min and 85 min separation gradients could be identified under the 130 min separation gradient， and 563 proteins could be exclusively identified， approximately 3.8 times more than those individually obtained under the 130 min separation gradient， which was attributed to the fact that the better separation of complex peptide samples and higher protein identification coverage could be achieved by the long separation gradient. Based on the above-mentioned merits， label-free quantitative analysis of the proteins from HCC patients and healthy individuals has been investigated. The total 2 062 protein groups were successfully quantified. Using P < 0.05 (T test) and 1.5-fold differences as the cutoffs， 56 proteins were considered to be differentially expressed， of which 33 were up-regulated and 23 was down-regulated in human plasma from HCC patients. Among them， as high as 69.6% of the significantly regulated proteins， such as CA1， LBP， TTR， and VTN， etc.， were reported as HCC associated proteins. These results were highly consistent with the results reported in the literature. In addition， the other differential proteins were not previously reported as HCC indicators and were found to have significant association with liver cirrhosis， fatty liver， and liver fibrosis. Bioinformatics analysis indicated that the differential proteins were involved in functions and processes closely related to the occurrence and development of HCC， such as serine-type endopeptidase inhibitor activity， haptoglobin binding action， immune response， and lipid transport， etc. The results indicates that the differential proteins may play a crucial role in the development of HCC， further research is ongoing to explore the potential of these proteins as biomarkers. [Conclusions] The standardized plasma proteomics workflow established in this study not only provides a valuable methodological framework for subsequent investigations in this field but also offers diverse technical solutions for public research platforms conducting plasma proteome profiling. The quantitative results of the plasma protein profile obtained in the study provide data support for the subsequent research on the mechanism of HCC development and potential biomarkers for the disease.
Laboratory Safety Classification and Grading Based on Protection Level and Random Forest Algorithm
Hu Xinjie;Xia Qi;Liu Zhe;Song Xiaofei;Yan Jin;Liu Yuhuan;[Objective] The classification and grading of laboratory safety is a fundamental prerequisite for safety management in higher education institutions. In recent years， this aspect has gradually become a focal point in the safety inspection work conducted by the Ministry of Education in China. Currently， research on laboratory safety classification and grading primarily concentrates on a single dimension of hazard sources， specifically utilizing hazard sources and their quantities to evaluate risks and determine safety levels. However， the risk level of laboratories is essentially a composite function of both the probability of hazard source release and protective efficacy. Recent laboratory accidents have also indicated that incidents do not typically occur in the most hazardous laboratories but rather in those with the poorest protective measures. Therefore， constructing a dual-dimensional evaluation model based on "hazard sources - protective levels" holds significant theoretical value and practical implications. [Methods] The study adopts big data analytics based on the random forest algorithm as the final risk evaluation method. The research data consists of two parts: 19 hazard source parameters and 7 protection level parameters.Gradually optimize the classification method through the following three comparative methods: (1) the gauge evaluation method that only considers the hazard source; (2) the expert evaluation method based on the level of protection， where the full-time safety supervisor provides the classification results based on the protection level parameters and hazard source parameters; (3) the big data method based on the random forest， which uses the classification results from the expert evaluation method as the original data and obtains the results after modeling according to the random forest algorithm. Compare the three sets of data twice. Compare the gauge evaluation method with the expert evaluation method to study the changes in classification results based on the protection level parameters. Compare the expert evaluation method with the big data method to investigate whether the big data method based on the random forest algorithm can effectively replace the expert evaluation method. [Results]The grading results of the regulatory evaluation method and the expert assessment method show significant differences. The expert assessment method considers safety supporting facilities， relevant regulations and systems， and safety training and drills as core indicators for risk evaluation. Laboratories with fewer hazards exhibit an upward trend in risk levels due to incomplete protective systems， while laboratories with a high concentration of hazards， which generally prioritize the development of protective measures， show relatively lower risk leve. This leads to a concentration of safety evaluation results in the moderate risk category. Optimized big data methods have demonstrated the feasibility to replace traditional expert evaluation methods， achieving a classification accuracy of 95.7% and grading accuracy of 94.8%. [Conclusions](1) The expert evaluation method based on protection levels is more aligned with the actual needs of laboratory safety management compared to the rating scale evaluation method. Its risk grading results show a significant correlation with the concentration distribution pattern of potential hazards， aiding in precisely identifying high-risk laboratories that require enhanced supervision.(2) The big data method constructed using the random forest algorithm has demonstrated technical feasibility to replace expert evaluations. The accompanying graphical user interface (GUI) software overcomes professional barriers， enabling personnel from non-safety domains and computer professionals to perform classification and grading operations at any time. This approach ensures the scientific nature of evaluations while significantly improving work efficiency.(3) The future development directions for laboratory safety classification and grading work include an integrated information management platform， 3D visualization technology， efficient workflows， and fine-grained management.
Solving Model and Experiment for Large-Scale Traveling Salesman Problems Based on Cluster Strategies
JIAO Dongbin;CHEN Zhiqing;WANG Fanghua;LI Ying;YANG Weibo;YAN Shi;[Objective]Integrating theory and practice is essential for training proficient AI professionals within the Emerging Engineering Education (3E) framework. This AI experiment employs the large-scale Traveling Salesman Problem (TSP)， which is a classic NP-hard challenge demonstrating the impracticality of exact methods at scale and the inconsistency of heuristic solutions， to connect abstract AI principles with hands-on implementation， offering a valuable learning opportunity for AI students. To address the high complexity， slow convergence， and poor scalability of existing approaches when applied to TSP instances with thousands of nodes， a novel cluster-based TSP framework， termed ClusterTSP， is proposed. [Methods]The proposed methodology strategically decomposes large-scale TSP instances into smaller subproblems via a clustering algorithm. This decomposition significantly reduces the computational burden compared to solving a single large instance.This "divide and conquer" strategy， aligning with efficient algorithm design principles， enables sophisticated optimization of the resulting subproblems. Following decomposition， ClusterTSP utilizes the representational power of deep learning， specifically employing Pointerformer， a cutting-edge method with notable TSP success， to optimize these smaller problems.Leveraging the Transformer network and its attention mechanisms， Pointerformer efficiently learns complex sequential dependencies to generate high-quality TSP tours. Applying it to clustered subproblems harnesses deep learning's ability to identify near-optimal local routes. This integration within a decomposition framework mitigates scalability issues common in end-to-end deep learning for very large TSPs. To construct a high-quality final tour that encompasses all the nodes of the original TSP instance， ClusterTSP incorporates a carefully orchestrated sequence of post-processing algorithms. These algorithms are designed to effectively merge the optimized subproblem solutions and further refine the resulting global path. Initially， an efficient multi-entry/exit greedy algorithm is employed to construct an initial global tour by intelligently connecting the optimized sub-tours generated by Pointerformer. This greedy approach provides a computationally inexpensive yet reasonably good starting point for subsequent optimization. Subsequently， a cluster boundary optimization algorithm is applied to specifically address the connections between nodes residing in different clusters， aiming to smooth transitions and eliminate potential suboptimal edges introduced by the initial decomposition. Finally， a 2-opt local search algorithm， a widely recognized and effective local optimization technique for the TSP， is implemented to iteratively improve the overall tour quality by systematically exploring small perturbations to the current solution and accepting those that lead to a shorter path. The synergistic interplay of these post-processing algorithms ensures enhanced solution quality by combining Pointerformer's localized optimization with the global coherence enforced by refinement steps. [Results]Extensive experiments on large-scale TSP instances (node size ≥103) demonstrate that ClusterTSP exhibits significant advantages in accuracy， efficiency， and scalability compared to the state-of-the-art deep learning benchmarks， including ACO， DRL＿PtrNet and Pointerformer.This advantage in solution quality becomes more pronounced with increasing problem size， and ClusterTSP effectively addresses the scalability limitations of traditional deep learning methods for large-scale scenarios. [Conclusions]Beyond its technical contributions， this research adopts a research-led teaching initiative， actively engaging students in the entire algorithm development and evaluation process.This hands-on experience fosters a deeper understanding of theoretical concepts， enhances practical skills， and significantly improves students' independent problem-solving abilities， exemplifying the "3E" initiative's emphasis on integrating theory and practice in AI education within an experimental teaching context.
A Study on the Comprehensive Performance Evaluation of University Laboratories under the Drive of Informatization
Shang Lei;Wang Haijie;Yin Ruitao;He Rui;[Objective] The objective of this study is to address the critical issues faced by university laboratories， which are central to academic research and educational activities. The main challenges are the low level of informatization in performance evaluation， frequent safety incidents， and uneven resource distribution. These issues not only hinder the laboratories' operational efficiency but also pose significant risks to the safety of faculty， students， and the integrity of research outcomes. The importance of this study lies in its potential to enhance the management and operational efficiency of laboratories， thereby fostering a safer and more productive research environment that is capable of meeting the demands of modern academic research. [Methods] The development of the performance evaluation index system is based on an integrated laboratory data platform. This system employs the Fuzzy Analytic Hierarchy Process-Entropy Weight Method (FAHP-EWM) to determine the weights of various performance indicators. This method is chosen for its ability to integrate expert opinions with objective entropy measures， ensuring a balanced assessment. Experts evaluate the importance of different performance indicators， and entropy is used to measure the amount of information each indicator provides， helping to identify which indicators are most informative. This dual approach allows for a more nuanced understanding of each indicator's contribution to overall performance. The Fuzzy Comprehensive Evaluation approach is then applied to assess the overall performance of the laboratories. This approach involves defining evaluation criteria and membership functions for each criterion， converting performance data into fuzzy values， and aggregating these values to provide an overall performance score. This method is particularly useful for handling uncertainty and imprecision in performance data， which is common in real-world scenarios. It allows for a more flexible and comprehensive evaluation that can adapt to varying conditions and data quality. Additionally， input-output analysis is integrated to assess the operational efficiency of laboratories. This involves analyzing the inputs (resources， time， etc.) and outputs (research outcomes， publications， etc.) of the laboratories to determine their efficiency. This analysis reveals underutilized resources and inefficiencies， providing clear directions for process optimization. This method provides a quantitative measure of how well the laboratories are converting inputs into valuable outputs， which is crucial for optimizing resource use and improving overall performance. The system is designed to be scalable and adaptable， making it suitable for laboratories of different sizes and with varying research focuses. This adaptability ensures that the system can be tailored to the specific needs of each laboratory， enhancing its applicability and effectiveness. The integration of these methods into a single system allows for a comprehensive evaluation that considers multiple aspects of laboratory performance， from safety and resource management to research output. [Results] The empirical results from the implementation of this system demonstrate significant improvements in evaluation informatization. There is a notable reduction in the probability of safety accidents， and enhanced protection of national assets and the safety of faculty and students. The system also contributes to the optimization of resource allocation， ensuring that resources are used efficiently and effectively， which is crucial for the long-term sustainability of laboratories. These results highlight the system's ability to provide innovative insights into the management of university laboratories， offering a scientific， standardized， and efficient approach to laboratory management. [Conclusions] This study presents a robust and comprehensive framework for evaluating and enhancing the performance of university laboratories. The integration of advanced analytical techniques with practical management strategies has resulted in a system that is both scientifically sound and operationally feasible. The findings and methodologies presented here are expected to serve as a valuable reference for academic institutions and researchers worldwide， promoting the continuous improvement of laboratory management practices. By providing a structured and systematic approach to laboratory performance evaluation， this study aims to facilitate the transition towards more efficient， safe， and sustainable laboratory practices， ultimately contributing to the advancement of knowledge and innovation in the academic sphere.
Strategies for the development and construction of laboratories under China-U.S. technological rivalry
SUN Jingyong;ZHAO Xiaofeng;WANG Jiani;MO Xiaoke;JIANG Shan;[Objective] Against the backdrop of China-US technological rivalry， US export control policies severely constrain laboratory construction and scientific collaboration within Chinese universities. This paper comprehensively analyzes the specific difficulties faced by targeted universities in equipment procurement and research cooperation， focusing on challenges such as restrictions on high-end equipment procurement and technology transfer， supply chain disruptions and soaring costs， severed international collaboration channels， and interrupted research projects. [Methods] Based on complex systems theory and incorporating collaborative innovation practices， the study proposes systematic countermeasures. Research indicates that breaking through technological barriers through independent R&D， establishing diversified procurement networks， and promoting equipment open access and revitalizing existing resources can effectively alleviate equipment acquisition challenges. Expanding collaboration through non-US channels and "organized research" can mitigate the impact of restricted international cooperation. The integrated countermeasures proposed provide actionable pathways for university laboratories under technological controls to overcome technological blockades and ensure efficient research operations. They also offer theoretical support and decision-making references for enhancing national technological self-reliance and building a self-sufficient scientific and technological innovation system. [Results] Laboratory construction and management encompass multiple facets including equipment procurement， resource allocation， team collaboration， and international cooperation. It constitutes a complex system composed of "hardware resources - software environment - collaborative actors." According to complex systems theory， the overall efficacy of such systems depends on the quality of interactions between components rather than the performance of any single element. Its nonlinear characteristics and emergent properties necessitate designing countermeasures from a systemic synergy perspective. Amidst the China-US technological rivalry， laboratory construction faces multifaceted challenges such as obstacles in procuring high-end instruments， difficulties in open access/sharing， and disruptions to international exchanges. Due to functional boundaries and resource limitations， any single entity finds it difficult to comprehensively resolve these systemic problems. Therefore， multi-actor collaboration is essential， bringing together the strengths of equipment management departments， research teams， cross-institutional platforms， and international exchange offices. Only through cross-domain coordination and resource integration can laboratory construction break through blockades and operate efficiently. [Conclusions] Under the constraints of China-US technological controls and embargoes， laboratory construction and management must adapt to the new environment. By enhancing indigenous R&D and production capabilities， optimizing procurement channels， strengthening collaborative sharing systems， and broadening cooperation avenues， laboratories can effectively counter challenges arising from restricted procurement of high-end equipment and constrained research organization. This approach improves resource utilization and research efficiency， ensuring high-functioning laboratory operations. Facing the intensifying US technological controls and embargoes against China， domestic universities must fully recognize the severity and urgency of the situation. Applying complex systems theory， early assessment and prevention are crucial. Laboratories should， based on their specific construction goals and development needs， identify lists of urgently required equipment， components， and materials. Procurement planning should be implemented step-by-step， considering urgency， procurement cycles， and the availability of alternative solutions. More importantly， a series of comprehensive national-level strategies must be formulated to safeguard national technological security and indigenous innovation capacity.
Construction Practice and Development Path of China’s University Humanities and Social Science Laboratory in the Era of Mega-Science
QU Liaojian;GUAN Yuanni;[Objective]In the mega-science era， real-world problems faced by countries worldwide are characterized by complexity and uncertainty， requiring interdisciplinary collaboration among experts from multiple fields. Humanities and social sciences seek new development through interdisciplinary cooperation with other disciplines， and some top international universities have initiated the construction of laboratories for humanities and social sciences. China has seized the historical opportunity of data-driven paradigm transformation in the mega-science era to advance the “new liberal arts” construction， promoting the innovative development of humanities and social science laboratories in universities. Efforts have been continuously made in reforming research paradigms of humanities and social sciences， strengthening international exchanges and cooperation， cultivating talents with both national sentiment and global vision， and producing research results that meet social needs. Although humanities and social science laboratories in China’s universities have achieved certain results over decades of construction， challenges remain in institutional support， data acquisition， talent cultivation， and cross-border integration. Current academic research rarely proposes solutions by drawing on international experience based on the historical positioning of the mega-science era. This study， rooted in the mega-science era， examines the environment and significance of constructing such laboratories in China， reviews their construction history since the founding of the People’s Republic of China， analyzes current dilemmas and challenges， and proposes optimization paths from the perspectives of ecological environment， university organization， and laboratory.[Methods]This study primarily employs data analysis， case analysis， and comparative analysis. First， based on the background of mega-science and the global revival of humanities and social sciences， it analyzes the necessity of constructing “new liberal arts” laboratories in China to meet the needs of the times and society. Second， it organizes relevant collected data， takes several humanities and social science laboratories as typical cases， and reviews the development history of humanities and social science laboratories in China’s universities. Then， by analyzing the history and current situation， it identifies the problems and challenges faced by these laboratories. Finally， from the perspectives of internationalization and interdisciplinarity， it draws on the excellent experiences of top international universities to propose development suggestions for China's university laboratories in this field.[Results]Through an in-depth analysis of the practical explorations of humanities and social science laboratories in China’s universities， this study argues that to address the dilemmas of imperfect institutional support， inconsistent overall planning， insufficient data resources， inadequate talent cultivation， ineffective interdisciplinary integration， and unsound organizational construction， development paths can be optimized from three levels:①At the ecological environment level， create a practical space with policy support， data assistance， and collaborative education.②At the university organizational level， improve a support system for resource integration， funding consolidation， and barrier-breaking.③At the laboratory level， explore a growth path of consolidating humanities and social science foundations， introducing and cultivating cross-border talents， and constructing laboratory matrices.[Conclusions]In the mega-science era， humanities and social science laboratories in China’s universities face new interdisciplinary research topics and a historical opportunity for data-driven paradigm transformation. Exploring the decades-long construction and development of humanities and social science laboratories in China’s universities reveals that from the founding of the People’s Republic of China to the pre-reform period， organized scientific research forces were accumulated; after the reform and opening-up， they were integrated and reconstructed on the original basis to actively serve national construction; subsequently， guided by national strategies， systematic transformation began， focusing on cultivating applied and interdisciplinary “new liberal arts” talents. These laboratories have always served as important platforms for constructing disciplinary， academic， and discourse systems of philosophy and social sciences with Chinese characteristics. In the torrent of profound social changes in the mega-science era， humanities and social science laboratories in China’s universities are facing challenges in institutional support and overall planning， data acquisition and talent cultivation， as well as interdisciplinary integration and organizational construction. To effectively address these challenges， urgent actions are required from the levels of ecological environment， university organization， and laboratory to optimize their development paths.
Intelligent monitoring system for hazardous sources in university laboratories based on IoT-enabled smart sensing
GU Wenyuan;YANG Fanfan;LI Wenwu;WANG Zhifei;ZHU Zhen;[Objective] University laboratories， vital hubs for talent development and scientific research， inherently contain diverse hazards—flammable chemicals， high pressure， radiation， and complex equipment. Traditional safety management relying on manual inspections and isolated monitoring systems suffers from delayed warnings， coverage gaps， and inefficient response coordination， leading to persistent accident risks. Motivated by national policies advocating smart research infrastructure and the urgent need for proactive safety governance， this research designed and implemented an intelligent hazard monitoring system. Its core objective is to transform lab safety from passive reaction to dynamic prevention by enabling real-time， comprehensive risk perception， intelligent early warning， and multi-level coordinated response， thereby safeguarding personnel and assets. [Methods] The system employs a meticulously designed four-layer IoT architecture: (1) Hardware Perception Layer: Diverse sensors (electrochemical/NDIR/PID gas detectors， temperature/humidity sensors， dust monitors， AI cameras with behavior recognition algorithms) are strategically deployed across labs， including concealed spaces like fume hoods， to capture real-time environmental and visual data. (2) Data Transmission Layer: A dual-channel strategy ensures stability: sensor data is structured via serial servers and stored in MySQL， while AI camera streams are encoded (H.264/AAC) and transmitted via RTMP/FLV to a streaming server (HLS protocol)， preventing interference in complex electromagnetic environments. (3) Platform Service Layer (Core Intelligence): Built on Spring Cloud Alibaba (Java)， this microservice-based layer integrates MySQL for structured data， Redis for caching， and the Camunda workflow engine. Key innovations include: Risk-based Dynamic Thresholds & Response: Lab classification (Levels I-IV based on hazard severity/type) dynamically adjusts alarm thresholds. Intelligent Analysis: Real-time fusion of multi-source sensor data and video analytics detects anomalies (micro-leaks， abnormal behavior， parameter coupling risks). Automated Multi-level Workflow: Upon alarm， Camunda triggers level-matched responses (local sound/light alerts + SMS/WeChat/call notifications escalating from lab personnel to school administrators) and retrieves relevant digital emergency procedures. (4) Application Interaction Layer: Vue 3/Element Plus (PC) and uni-app /uView (Mobile) interfaces provide unified visualization of real-time data， video feeds， alarms， device status， and facilitate emergency process participation. Three core functional modules synergize on this architecture: (1) Lab Classification Engine: Automates risk assessment (I-IV) and categorization (Chemical， Biological， etc.) based on hazards， providing the foundation for differentiated monitoring policies. (2) Intelligent Monitoring & Warning: Leverages classification to deploy tailored sensor thresholds and AI video rules. Detects breaches and triggers the hierarchical alert/response workflow. Integrates a digital emergency procedure library. (3) Data Visualization Dashboard: Offers comprehensive "cockpit" views (real-time video， risk distribution maps， alarm stats， device health) for multi-dimensional analysis and decision support at room， building， and campus levels. [Results] Deployed across 249 high-risk labs (Levels I & II， including those handling explosive/toxic chemicals) at Xi'an Jiaotong University， the system demonstrated significant impact: (1) Enhanced Efficiency & Response: Automated classification and monitoring reduced safety inspection time by over 90%. A critical real-world incident validated effectiveness: At 3:27 AM， the system detected a CO micro-leak (58ppm) in a Level II Chem lab. It instantly triggered Level 2 alerts (local alarms + multi-level notifications). Personnel confirmed remotely via video， responded， and resolved the leak (caused by an unsecured valve) within 26 minutes， preventing potential explosion/poisoning. Post-analysis confirmed the leak would likely have gone undetected for hours without the system. (2) Robust 24/7 Protection: A network of 780 gas detectors， 465 AI cameras， and 412 environmental sensors provided continuous coverage. User surveys (n≈300) indicated 96.2% of staff/students felt significantly safer， particularly citing reliability during unmanned hours. (3) Optimized Coordination: The system automated over 2，400 lab info updates and facilitated seamless "Lab-Department-University" coordination. Of 1，482 valid alarms， 21% required departmental intervention and 3.9% required university-level response， proving the hierarchical workflow. (4) High Accuracy & Actionable Insights: System accuracy reached 96.61% (1，482 valid alarms out of 1，534 total). Primary triggers were Humidity (67.27%， mainly summer highs prompting dehumidification)， VOC (21.12%， exposing procedural lapses like spills/poor ventilation)， and Dust (4.59%， driving improved dust control). Alarm data is actively used to refine thresholds (e.g.， seasonal humidity adjustments， gas thresholds by lab level) and provides a quantitative foundation for EHS system development. [Conclusions] This IoT-based intelligent monitoring system represents a significant paradigm shift in university laboratory safety management. By architecting a synergistic framework integrating multi-source sensing， risk-based dynamic response， and data-driven visualization， it achieves real-time hazard perception， precise early warning， and efficient closed-loop incident management. Practical deployment confirmed substantial improvements in safety outcomes， operational efficiency， and collaborative emergency response. While challenges like sensor reliability in extreme environments and AI adaptability to novel scenarios require ongoing mitigation (e.g.， multi-sensor fusion， edge computing， federated learning)， the system provides a scalable， proactive safety model. Its "Perception-Decision-Response" mechanism offers a valuable reference for lab safety standards. Future integration with access control， energy management， and campus security platforms promises a holistic safety ecosystem underpinning high-quality scientific research.
Safety Management Exploration of Large-scale Self-made Equipment in University Laboratories
LIU Zhao;ZHU Zhengmao;[Objective] The study aims to investigate the status， existing problems， and management strategies of large-scale self-made equipment in university laboratories， emphasizing safety management throughout the entire lifecycle of such equipment. It also showcases the practical experiences from North China Electric Power University， highlighting the necessity for systematic and standardized safety management practices. [Methods] By analyzing the characteristics and management processes of large-scale self-made equipment， this paper identifies key stages where safety risks may arise， including design and manufacturing， installation and commissioning， operation and usage， and scrapping. The research methodology includes literature review， comparative analysis between self-made and commercial off-the-shelf (COTS) equipment， and case studies focusing on the safety management practices at North China Electric Power University. Emphasis is placed on the need for comprehensive safety assessments， risk control measures， and the development of specific emergency response plans tailored to the unique aspects of self-made equipment. [Results] The findings reveal that large-scale self-made equipment， due to its high customization， complexity， and innovation， poses significant safety challenges across all lifecycle phases. Key issues identified include inadequate consideration of safety and environmental factors during the design phase， lack of professional oversight during manufacturing， insufficient operator training， unclear safety responsibilities， and absence of clear guidelines for safe disposal. North China Electric Power University's approach demonstrates effective strategies for addressing these challenges， such as establishing a robust safety management system， implementing strict supervision during the manufacturing process， conducting thorough safety evaluations before installation， ensuring proper training for operators， and developing detailed standard operating procedures (SOPs).[Conclusion] Effective safety management of large-scale self-made equipment requires a holistic lifecycle perspective， integrating rigorous risk assessment and control measures at each stage. Universities should develop comprehensive safety management systems tailored to their specific needs， incorporating elements like safety standards， regular inspections， and emergency preparedness. The case study of North China Electric Power University illustrates how adopting a proactive stance towards safety can significantly enhance the safety and efficiency of laboratory operations involving self-made equipment. Furthermore， leveraging information technology to support the management of self-made equipment could further improve management effectiveness and reliability. This study underscores the importance of continuous improvement and adaptation in safety management practices to meet evolving safety standards and technological advancements. The paper concludes by advocating for the establishment of a unified set of safety management norms and technical standards for large-scale self-made equipment in universities， emphasizing the significance of interdisciplinary collaboration， regulatory compliance， and ongoing education and training for personnel involved in the lifecycle management of such equipment. Through these efforts， it is hoped that higher education institutions will be better equipped to handle the unique challenges posed by self-made equipment， thereby safeguarding both human lives and scientific endeavors.
Model Experimental Study on the Interaction Mechanical Characteristics of Tunnels First then Metro Station
FENG Jimeng;LI Yifei;SONG Jiadai;ZHANG Junru;The "Tunnels First， Station Later" (TBS) construction method is increasingly adopted in urban core areas with upper-soft and lower-hard strata to resolve scheduling conflicts in metro projects. However， the interaction mechanism between shield tunneling and subsequent mined station excavation remains poorly understood， posing significant risks to ground stability and structural safety. This study， based on the Guangzhou Metro Memorial Hall Station project， aims to clarify how the timing of shield passage—specifically before or after the construction of the station’s arch support—affects the mechanical behavior of the surrounding rock and supporting structures， providing a theoretical and experimental basis for optimizing construction sequences under similar complex geological conditions. [Methods] A combined approach of physical model tests and numerical simulations was employed. A geomechanical model test system was designed following standard similarity relationships， utilizing stratified similar materials with differentiated mix proportions to accurately simulate the mechanical characteristics of the upper-soft and lower-hard composite strata. Two representative working conditions were established: Condition 1， where shield tunneling takes place after the arch support is installed， and Condition 2， where shield tunneling occurs before arch support formation. Comprehensive monitoring covered surface settlement， subsurface displacement， and stresses in shield segments and the arch support. [Results] 1) Surface settlement during PBA station construction followed a typical unimodal Peck curve. Shield tunneling conducted after arch support formation (Condition 1) resulted in a maximum settlement of 42.1 mm， which was 5.9% lower than the 44.6 mm observed in Condition 2. This indicates that a completed arch support significantly reduces surrounding rock disturbance and enhances settlement control. 2) The stress on shield segments was notably lower in Condition 1， with a maximum principal stress of 1.31 MPa， compared to 1.63 MPa in Condition 2—an increase of 24% in the latter. Condition 2 also exhibited more extensive stress concentration areas and tensile regions on the inner surface of segments， indicating a less favorable mechanical state. The pre-existing arch support in Condition 1 allows for better load release and redistribution， reducing demands on the segments. 3) The arch support above the main station excavation bore the highest structural stress， with a maximum compressive stress of 6.7 MPa in Condition 1， significantly higher than the 4.7 MPa in Condition 2. This confirms its role as the primary load-bearing component. Subsequent shield tunneling had limited impact on the stress state of the arch support in both conditions. Validation showed excellent agreement between model tests and numerical simulations， with errors in key metrics below 6%， confirming the reliability of the findings. [Conclusions] Both physical model tests and numerical simulations demonstrate that the timing of shield tunneling relative to arch support construction is critical in the interaction mechanics of "Tunnels First， Station Later" projects in upper-soft and lower-hard strata. Prioritizing the completion of the arch support before shield passage (Condition 1) provides superior settlement control and reduces shield segment stresses， making it the recommended sequence for enhanced safety and performance. If shield tunneling must precede arch support (Condition 2)， additional measures such as ground improvement or segment reinforcement should be considered to mitigate the identified risks of increased settlement and adverse structural loading. The developed model test system and analytical methodology offer a reliable and adaptable framework for studying similar complex tunneling-station interaction problems in underground engineering.