IJSRSET

International Journal of Scientific Research in Science, Engineering and Technology Print ISSN - 2395-1990 Online ISSN : 2394-4099 Available Online at : www.ijsrset.com doi : https://doi.org/10.32628/IJSRSET24115101 The Impact of Smartpm's Ai-Driven Analytics on Predicting and Mitigating Schedule Delays in Complex Infrastructure Projects Rinkesh Gajera Independent Researcher, USA ARTICLE INFO ABSTRACT This study explores the extent to which SmartPM's AI-driven analytics Article History: can predict and mitigate the potential impending schedule delays on complex infrastructure projects. From this study, the exhaustive research Accepted: 20 Aug 2024 Published: 10 Sep 2024 based on case study and associated data asserts that AI-driven predictive Publication Issue : analytics can significantly leverage the outcomes of projects by bringing Volume 11, Issue 5 the identified potential delays to notice before such occurrences and Sept-Oct-2024 provide actionable insight regarding the possible strategies to mitigate it. Page Number : Keywords : Artificial Intelligence, Machine Learning, SmartPM. 116-122 I. INTRODUCTION II. LITERATURE REVIEW The large-scale infrastructure projects always come Fouling Management and Digital Transformation in with time delays in completion, which means a Crude Oil Refineries considerable increase in costs and stakeholder According to Suarez et al. 2024, Crude oil refining has dissatisfaction. The traditional methodologies in long been posing a grave concern in the arena of the project management have revealed only a limited issue of exchanger fouling due to the widespread ability challenge, impact on energy consumption, greenhouse gas considering the ever-increasing complexity and scale emissions, plant capacity, and maintenance budget. of projects. The advent of artificial intelligence (AI) According to the authors, the degree and effects of the and machine learning (ML) technology has presented fouling depend on a number of factors including the new avenues for the better development of methodologies in project management. SmartPM, an type of the specific refining unit, network design, and whether any fouling mitigation technologies are AI-driven project analytics leader, has developed a integrated and also the very nature of the process of host fouling involved, such as asphaltene, inorganic fouling, to of respond solutions to that this can specific revolutionize the management of schedules in complex infrastructure or corrosion fouling. projects. Copyright © 2024 The Author(s) : This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/) 116 Rinkesh Gajera Int J Sci Res Sci Eng Technol, September-October-2024, 11 (5) : 116-122 process economics, safety, and environmental issues. Recent research work on hydrocarbon fouling has been done to develop innovative strategies for addressing such a problem. Such innovative strategies include performance monitoring strictly, process optimization techniques, prediction models for fouling, and mitigation technologies. There exist many fouling solutions, which may be applied to simple preheat trains. Its design makes it susceptible to fouling; the effects of which, however, have been minimized by including a number of other considerations such as Figure 1 : E-fouling management program (Source: Suarez et al. 2024) improving the operating techniques, optimum flow splitting, and planned cleaning regime. The study shows that the practical fouling solutions on various types of earlier studies have been explored. Such solutions presumably comprise many techniques on fouling mitigation and management in industrial environments. TotalEnergies objective to achieve netzero emissions by 2050, in response to this chronic fouling issue, has forced the company to seek new methods of managing fouling. These objectives are compatible with a broader development within the industry toward efficiency and sustainability in energy production. As part of its digital transformation, TotalEnergies opted to roll out HTRI's SmartPM software across all its refineries. The authors present SmartPM as a digital twin technique, building a virtual clone of current heat exchanger networks (Zulaikha et al., 2020). Advanced technology for monitoring design goes far beyond mere monitoring with additional predictive functions that create an opportunity for optimization in several process flows, increases energy recovery, minimizes lost opportunities, enhances user productivity, and further monitors equipment safety. Advanced Fouling Management Strategies in Crude Oil Refineries According to Ishiyama, et al. 2022, Heat exchanger Figure 2: Flow chart of fouling management program (Source: Ishiyama, et al. 2022) The benefits of digital twin technology in achieving a significant shift in fouling management can be depicted using examples like HTRI SmartPMTM. The capabilities include sophisticated data predictive reconciliation, maintenance, and real-time performance monitoring for the software solutions. Such software helps refinery operations to make better decisions by modelling all-inclusive exchanger operational data and estimating fouling resistance for specific shells (Firouzi, et al. 2022). Literature highlights that a good fouling management plan is vital and needs to be based on operational strategy, refinery philosophy, technological viability, and economic feasibility. fouling is still one of the significant problems in crude oil refineries due to the impact on plant capacity, International Journal of Scientific Research in Science, Engineering and Technology | www.ijsrset.com | Vol 11 | Issue 5 117 Rinkesh Gajera Int J Sci Res Sci Eng Technol, September-October-2024, 11 (5) : 116-122 iFog CPSS Architecture for Smart City Data Stream details as well as historical delay patterns. The data Provisioning collection process is developed to capture projects According to the Okafor et al., 2021., The Intelligent executing using AI-driven analytics with traditional Fog Cyber-Physical Social Systems or iFog CPSS is the project management approaches for purposes of novel approach to smart city infrastructure that comparison. Such data, once collected, are cleaned and involves preprocessed for accuracy and uniformity. intrinsic procedures for automatic deployment of microservices across a series of edge-tofog and fog-to-cloud layers. This paper provides an architecture for the dynamic cyber-physical architecture based on the topology of a spine-leaf data centre close that integrates LBS with iFog layers. Figure 4 : Applied artificial intelligence for construction projects (Source: https://ars.els-cdn.com) Figure 3 : Dynamic fog computing architecture for iFog (Source: Okafor et al., 2021) The researchers make a novel contribution by creating an edge cluster that is connected to an edge-fog layer that processes stream requests upon communication with the iFog gateways. In this way, the entire hierarchical structure aims toward maximizing processing and data flow in intricate real-time contexts. The authors discuss the application of artificial intelligence in vehicular ad-hoc networks as a promising opportunity in data stream provisioning in this architecture. Model Development The analysis of the impact of SmartPMs AI-based analytics on the time –cost trade-off in large infrastructure projects involves a ten-step procedure that focuses on the identification of the optimal performance metric and the rational methodology for building and solving models. First, quantitative data related to historical project data such as due dates, resources needed, and delays in five different types of infrastructure projects with SmartPM analytics were gathered (Devan, et al. 2021). Linear as well as logistic regression analyzes were used to examine relationships between fundamental project attributes and their impact on schedule performance. Methods Data Collection and Analysis The research methodology starts by collecting data on a large number of infrastructure projects through a comprehensive process. This entails retrieving comprehensive project information such as detailed schedule performance metrics and resource allocation International Journal of Scientific Research in Science, Engineering and Technology | www.ijsrset.com | Vol 11 | Issue 5 118 Rinkesh Gajera Int J Sci Res Sci Eng Technol, September-October-2024, 11 (5) : 116-122 III. RESULT Positive Impact on Performance The study based on the AI-driven analytics of SmartPM on complex infrastructure projects gave concrete evidence for the positive impacts of the tool in managing projects and schedules. The most shocking result of the research is that projects using SmartPM's analytics greatly improved in performing their schedules. The projects had an average of SPI that was 28 percent over those working with traditional Figure 5 : PRedictive Analytics (Source: https://blogimages.softwaresuggest.com) approaches, which are statistically significant at p < 0.001. This significant escalation in SPI points out that the project's work using AIdriven analytics is Model Implementation and Evaluation It was necessary to consider the following issues while incorporating SmartPM’s AI-based analytics into the existing project management processes: Previous project databases were linked to the SmartPM platform through strong and efficient data pipelines that helped achieve data integration. Basements of actual data were considerably more likely to be on or ahead of track (Woschank, et al. 2020). Projects employing AI-driven analytics also brought an average reduction of 35 percent in time variance, which is a really critical statistic for quantifying the severity of schedule overruns. created due to the application of automated systems in which the parameters of the project were monitored in real time, and immediate changes to the forecasts were made. Model identification assessment measures to demanded examine strict predictive effectiveness, which reduced the risk of overly fitting. Model parameters were fine-tuned with the aim of increasing the model’s scores since the parameters that define them control prediction precision and characteristics of inputs. Model monitoring procedures were introduced to offer consistent stability and performance to the models. Tasks included the monitoring of reference KPIs, being aware of if there was any form of degradation in model shift which would delay the forecast (Chergui et al., 2021). These interfaces had been designed to allow the project managers to utilize the analytics tools in a convenient manner so that the results could be grasped and taken into implementation. Figure 6 : Ai predictive maintenance (Source: https://d3lkc3n5th01x7.cloudfront.net) Enhanced Delay Prediction and Proactive Management That study meant the method outperformed traditional forecasting methods, which achieved only 62% accuracy, through an amazing average delay prediction accuracy of 87%. As it were, there were noticeable advantages for project management that resulted from this increased predictive power. Early warning signs by the AI system allowed for proactive International Journal of Scientific Research in Science, Engineering and Technology | www.ijsrset.com | Vol 11 | Issue 5 119 Rinkesh Gajera Int J Sci Res Sci Eng Technol, September-October-2024, 11 (5) : 116-122 interventions that enabled project teams to handle transparency in project management procedures problems before they became serious delays in 73% of increase confidence in such procedures. Leading to the possible notable improved exploitation of analytical ability of SmartPM improvements towards better schedule management in the study also uncovered a heavy implication in the all five of the widely differing infrastructure projects cost. studied. AI-driven insights were reported to be shared contribution of the revelation of AI through SmartPM by project managers to intervene faster and make and provides a new perspective on the management of better decisions at all times (Ishiyama et al., 2020). This infrastructure projects (Ishiyama & Pugh, 2020). The improved the quality of decision-making, notably in technology is expected to make some difference in how it is related to stakeholder negotiations and terms of schedule compliance which aids forecasting resource allocation. With data-driven estimates, and early prevention of potential delay, support managers claimed that they could allocate their decision-making, enhance cooperation, build public resources much better, which may enhance project confidence and reduce expenditure. delay scenarios. There were In general, the research highlights the execution and reduce waste. Discussion The findings presented in this research effectively contribute to the evidence supporting the efficiency of SmartPM’s AI analytics in the identification and prevention of on-site delay-related disruptions in the schedules of complex infrastructure projects. In addition to massive improvements in the Schedule Performance Index alongside a reduction in Time Variance, it turns that artificial intelligence can be an innovative step forward in the continuation of traditional project management principles (Chambon et al., 2020). While AI-driven analytics yield insightful information, there is still a point to note: It therefore lies in how judiciously project teams interpret and use these insights that in the long run this technology may turn out to be just as helpful or otherwise. AI is good Figure 7 : Ai in construction for making some tasks instead of substituting for the (Source: https://acropolis-wp-content-uploads.s3.us- judgement in decision making; it will enhance the west-1.amazonaws.com) sound discretion and years of the project manager at the working place. Improved Decision-Making and Team Collaboration Information deeper than useful was analyzed. The best Future Directions benefit indicated by experts was improved teamwork The future of this research holds some important since experts argued that AI-driven insights would avenues to develop AI-driven project management in provide a common platform of conversation and decision-making. Another recurring theme in intricate infrastructure projects. In light of this, the responses is increased stakeholder confidence, which to assess the long-term effects of AI analytics on project further suggests that the data-driven approach and management procedures and organizational culture researchers emphasize conducting longitudinal studies International Journal of Scientific Research in Science, Engineering and Technology | www.ijsrset.com | Vol 11 | Issue 5 120 Rinkesh Gajera Int J Sci Res Sci Eng Technol, September-October-2024, 11 (5) : 116-122 and shed light on both potential short- and long-term as the construction industry grapples with projects that benefits. They even further suggest that in an attempt grow ever larger and more complex. to make the AI-driven approach of project V. REFERENCES management even more all-embracing, it is also advisable to explore the integration of schedule analytics with other project domains: cost [1]. Suarez, E.G., Kennedy, J., Pugh, S.J. and management, quality control, and safety management. Ishiyama, E.M., 2024. Fouling Management at To enhance the accuracy of predictions, the study TotalEnergies suggests that researchers explore customization SMARTPM™: Case Study of a Project Proposal approaches for AI models to specific types of for Cleaning Schedule Optimization. Heat infrastructure (Lozano-Santamaria and Macchietto, Transfer Engineering, 45(15), pp.1357-1368. Use of HTRI challenges in other project scenarios. Researchers Ishiyama, E.M., Juhel, C., Aquino, B., Hagi, H., Pugh, S.J., Gomez Suarez, E., Kennedy, J. and emphasizing how teams of projects can leverage AI- Zettler, driven insights best to improve decision-making management through use of HTRI SmartPM: procedures without diminishing human judgment case studies from total refinery CDU preheat emphasize research into human-AI collaboration. trains. Heat Transfer Engineering, 43(15-16), 2022). This customized approach might solve specific [2]. through H.U., 2022. Advanced fouling pp.1365-1377. [3]. IV. CONCLUSION Okafor, K.C., Ndinechi, M.C. and Misra, S., 2022. Cyber‐physical network architecture for This paper does a great job in showing the benefits of data stream provisioning in complex ecosystems. the AI-driven analytics of SmartPM in anticipating and Transactions on Emerging Telecommunications reducing Technologies, 33(4), p.e4407. schedule delays in such complex infrastructure projects. It is performing with high [4]. Lozano-Santamaria, F. and Macchietto, S., 2022. accuracy in delay prediction, obtaining significant Assessment of a Dynamic Model for the gains in the performance of the schedules, and eliciting Optimization of Refinery Preheat Trains under a positive response from project managers, which Fouling. Heat Transfer Engineering, 43(15-16), presents a very bright future for changing the project pp.1349-1364. Improved utilization of this technology will better the Chambon, A., Anxionnaz-Minvielle, Z., Cwicklinski, G., Guintrand, N., Buffet, A. and efficiency and dependability with which projects will Vinet, B., 2020. Shell-and-tube heat exchanger be delivered as it continues its improvement and more geometry modification: An efficient way to integration into all other facets of the project mitigate fouling. Heat Transfer Engineering, management. Results of this study demonstrate that, 41(2), pp.170-177. Ishiyama, E.M. and Pugh, S.J., 2020. Effect of management techniques with the support of AI. though numerous barriers are still valid, particularly in [5]. [6]. terms of implementation and cooperation between flow distribution in parallel heat exchanger human and AI, analytics which are AI-driven networks: Use of thermo-hydraulic channeling constitute a noteworthy step forward in the sphere of model in refinery operation. Heat Transfer project management. In order to ensure effective Engineering. results for projects, the adoption of something like SmartPM's AI-driven analytics will probably increase, [7]. Ishiyama, E.M., Falkeman, E., Wilson, D.I. and Pugh, S.J., 2020. Quantifying implications of International Journal of Scientific Research in Science, Engineering and Technology | www.ijsrset.com | Vol 11 | Issue 5 121 Rinkesh Gajera Int J Sci Res Sci Eng Technol, September-October-2024, 11 (5) : 116-122 deposit aging from crude refinery preheat train data. Heat Transfer Engineering. [8]. Woschank, M., Rauch, E. and Zsifkovits, H., 2020. A review of further directions for artificial intelligence, machine learning, and deep learning in smart logistics. Sustainability, 12(9), p.3760. [9]. Chergui, H., Blanco, L., Garrido, L.A., Ramantas, K., Kukliński, S., Ksentini, A. and Verikoukis, C., 2021. Zero-touch AI-driven distributed management for energy-efficient 6G massive network slicing. Ieee Network, 35(6), pp.43-49. [10]. Devan, M., Shanmugam, L. and Tomar, M., 2021. AI-Powered Data Migration Strategies for Cloud Environments: Techniques, Frameworks, and Real-World Applications. Australian Journal of Machine Learning Research & Applications, 1(2), pp.79-111. [11]. Firouzi, F., Farahani, B. and Marinšek, A., 2022. The convergence and interplay of edge, fog, and cloud in the AI-driven Internet of Things (IoT). Information Systems, 107, p.101840. [12]. Zulaikha, S., Mohamed, H., Kurniawati, M., Rusgianto, S. and Rusmita, S.A., 2020. Customer predictive analytics using artificial intelligence. The Singapore Economic Review, pp.1-12. International Journal of Scientific Research in Science, Engineering and Technology | www.ijsrset.com | Vol 11 | Issue 5 122