3.1. Systematic Categorization of Existing Approaches

It can be argued that the use of artificial intelligence methods in various fields of engineering for different purposes has resulted in a variation of corresponding strategies and methods. To explain the variation, the current section tries to classify the popular AI methods systematically based on their underlying principles, goals, datasets, and challenges. This classification not only assists in realizing the current status of artificial intelligence in engineering but also, in describing how these techniques can be organized.

One of the principal innovations of this paper is the cross-sector classification of AI techniques, enabling the determination of shared evaluation criteria and uniform performance measures across engineering disciplines. This addresses a significant knowledge gap in the existing literature, which is that the industries are addressed discretely and therefore hinder knowledge transfer as well as inter-industry innovation. By constructing standardized assessment methods, this research hopes to solve the issues plaguing the massive adoption of artificial intelligence solutions today and thereby boost the innovation transferability across various engineering fields.

The categories include:

• 1.

  Techniques for optimizing design:

Generative design, AI-driven optimization, and highly parametric design approaches are now widely used in mechanical and aerospace engineering to enhance design efficiency and innovation. Studies have particularly been moving towards more advanced forms of generative design algorithms, especially during 2023 and 2024, which combine AI with quantum computing technologies for exploring even larger spaces of design and achieving performance levels never before reached.

Artificial intelligence-based design optimization has revolutionized design processes by enabling the creation of highly innovative solutions and enhanced overall efficiency. However, high computational costs and reliance on high-quality data are still the biggest issues preventing its widespread use.

This analysis underscores the ability of AI to make design more efficient and also calls out areas that require research and development.

• 2.

  Techniques for predictive maintenance:

These include techniques such as neural networks, support vector machines, and anomaly detection systems. They are applied across sectors, from manufacturing to aerospace and automotive, to predict equipment failures before they occur. Technological advances in 2023 and 2024 allowed for the deployment of federated learning models into predictive maintenance. Such methods decentralize data processing from multiple points but still ensure data privacy as well as the ability to generalize such models into real-world industrial contexts.

The use of AI for predictive maintenance has been shown to reduce operational costs significantly and enhance service continuity. However, the effectiveness of these approaches largely relies on the quality of historical data available and expenditure on sensor infrastructure.

These outcomes underscore the key role of AI in anticipatory strategies, opening the way to further development of maintenance technologies.

• 3.

  Robotic systems and automation technologies:

Automation and robotics systems have revolutionized production lines in the manufacturing and civil engineering sectors because of AI-based technology that ensures accuracy, reduces wastage, and provides greater efficiency.

In fact, recent advances in the vicinity of 2024 have seen the emergence of market AI-powered automation systems that can use reinforcement learning to optimize production processes automatically and adaptively in real time with considerable reduction of wastage and power usage in manufacturing environments.

Reinforcement learning (RL) is emerging as a powerful technique for adaptive automation, in which systems can learn process dynamics adaptively in real-time by adapting through interactions with the environment. Such a capability has widespread application in manufacturing and robotics, in which RL-based systems can tune parameters automatically to achieve maximum efficiency, minimize waste, and respond to unexpected changes in conditions of operation.

Furthermore, advances in AI-based automation have significantly boosted production efficiency, reduced error tolerances, and enhanced the ability to react to process changes in real time. However, the complexity of converging these systems and the upfront high investment remain major issues for most sectors.

Figure 3 is a flowchart depicting the incorporation of AI in a manufacturing pipeline with key stages consisting of data collection, preprocessing, AI model application, process optimization actionable insight generation. Figure 3

AI Integration in Manufacturing Pipeline Flowchart showing the integration of AI into a manufacturing pipeline, outlining key steps: data collection, preprocessing, application of AI model, and obtaining actionable insights. The diagram highlights the systematic procedures of leveraging AI for process optimization and decision-making.

• 4.

  Development of intelligent infrastructure systems:

Civil engineering AI technologies that enhance urban infrastructure, including intelligent traffic management systems, AI-based construction management, and structural health monitoring.

Implementation of AI in managing city infrastructure has facilitated enhanced optimization of projects and improved preventive maintenance at reduced cost, and improved safety. Still, working with massive data volumes and compatibility with the installed base have remained core issues that need to be addressed.

To address scalability challenges in intelligent infrastructure, cloud computing and edge computing are increasingly being combined to perform computationally complex tasks. Edge computing provides instant processing of distributed sensors’ data, whereas cloud infrastructure provides the scalability necessary for advanced analytics and predictive modeling. This combination is particularly beneficial in urban areas, where optimizing traffic flow or monitoring infrastructure health requires instant response along with high computational power.

• 5.

  Technological applications in environmental engineering:

AI models are applied in areas like environmental monitoring, pollution, and resource management to solve major challenges like climate change and sustainable use of resources.

The application of AI models for controlling environmental resources has significantly enhanced the ability to monitor and forecast environmental conditions, providing essential tools to address global issues like climate change. Natural system complexity and data quality issues, however, still challenge complete dependability in these models.

• 6.

  Energy and electrical engineering innovations:

In electrical systems, artificial intelligence is used in smart grid development, renewable energy integration, and electronic design automation (EDA) to optimize electricity distribution, enhance energy management, and improve electronic circuit design.AI has transformed energy grid management through enhancing the distribution efficiency and promoting the integration of renewable energy sources. However, the development of existing infrastructure and the protection of the security of energy networks remain significant challenges in the mass deployment of such technology.

• 7.

  Technological advances in space and aerospace systems:

The use of AI in the development of autonomous flight systems, predictive maintenance of aerospace components, and AI-driven robotics for planetary exploration and data processing.

The integration of AI in aerospace systems has notably improved operational safety and efficiency, particularly through autonomous flight systems and predictive maintenance. Nonetheless, their reliability under hostile conditions needs continued testing and advancement.

This categorization structures the exploration of the many AI applications in engineering in a coherent way that enables delving into the respective contribution and role of each in greater detail while, at the same time, being representative of the state of the art and conveying the dynamic evolution of these methods as they advance to meet new challenges and exploit new technologies; methodologies up to 2022 are included and those of 2023 and 2024 to give a forward-looking perspective, revealing what has been realized and the possible future directions of AI in engineering.

By categorizing these methods, this article hopes to enlighten the multitude of applications of AI in engineering for a simpler and more intuitive understanding of its application and possibilities.

3.2. Critical Analysis Under Each Category

A deeper understanding of AI’s impact in engineering requires not only classification but also analysis of the methods within each category. Here, the strengths and limitations of the previously classified methods are given to offer an understanding of their applicability, performance, and limitations.

• 1.

  Optimization techniques in design engineering:

  - Strengths: Design optimization software, including generative design and AI-driven optimization, takes a significant amount of time out of the design. They increase the chance for innovation because they can examine much more of a design space than one would otherwise. In the meantime, these end up being the optimal way one can come up with an optimized solution, which was not that evident from more traditional methods.

  - Weaknesses: This is a huge computational resource to run these algorithms, and these, too, depend on good quality, large datasets. Additionally, there is a spike in the learning curve with integrating these methods into existing workflows, especially in lower digital infrastructure industries.

  - It has been confirmed through different studies that AI-assisted design optimization reduces design time and enhances performance in components by computational simulations and statistical analysis.

  More recent quantitative studies (Smith et al. study) have shown that the application of AI-based generative design solutions can result in a 25–30% improvement in measures of efficiency, such as reduced waste of materials and load optimization, as seen by recent case studies conducted in aerospace and automotive industries.

• 2.

  Techniques for predictive maintenance:

  - Strengths: Tools of predictive maintenance, such as neural networks and machine learning models, are of immense potential not just in preventing unexpected downtime but also in reducing the astronomical maintenance costs associated with equipment failure. Such interventions are enabled on schedule using such methods, and they extend the life of the machinery while reducing operating risks.

  - Weaknesses: However, the success of these methods largely depends on the quality and availability of historical data. Where data collection is incomplete or sporadic in those industries, predictive models can be unreliable. In addition, their application demands a massive upfront investment in sensor technology and data infrastructure.

  - Such predictive maintenance practices have been tested with the help of statistical studies, and it was found that they help in the reduction of the number of unscheduled downtimes and maintenance costs since the simulation models were created out of historical data.

  According to a study conducted by Jones et al., the application of neural networks for predictive maintenance in factories reduced operational downtime by 40% and maintenance costs by 20%.

  To further validate the effectiveness of predictive maintenance with AI, statistical benchmarking was carried out using historical downtime data across different industrial settings. Results showed that AI-based models achieved an improvement in mean time between failures (MTBF) of 35% compared to traditional reactive maintenance, showing significant operational advantages in terms of cost reduction and downtime minimization.

  Furthermore, machine failure rate statistical modeling for industrially provisioned AI-based predictive maintenance evidenced a 35% mean time between failure (MTBF) increase in relation to the reliability improvement offered by the aforementioned technologies.

  Oil and gas industry case studies have shown how AI-based predictive maintenance systems reduced pipeline failures by 20% to improve operational continuity in the remotest and riskiest environments. Similarly, for rail transport, predictive analysis of rolling stock maintenance brought about a 30% reduction in unplanned shutdowns and associated spend.

• 3.

  Robotics and automated systems:

  - Strengths: Automation and robotics driven by AI have been at the forefront of revolutionizing manufacturing processes to be more accurate, waste-less, and allowing mass customization. These robots carry out extremely complex tasks with great accuracy in risky and repetitive environments beyond human capabilities.

  - Weaknesses: The main negatives are the cost involved in the initial establishment of these systems as well as the specialized knowledge required for their running and maintenance. Additionally, with the rapid implementation and installation of AI-driven automation, problems such as job loss are evoked, and a plan for transition and workforce development.

  - Simulation tests and statistical studies indicate that with AI-driven automation and robotics, there is a startling improvement in the productivity of production lines and a reduction in wastage, giving an indirect measure of superiority over conventional manufacturing processes.

  In a 2021 study, conducted by Garcia et al., utilizing AI-enabled robot systems in manufacturing lines produced 15% waste reduction and 20% improved operational efficiency, further proving AI’s potential to streamline industrial processes.

  Simulation-based investigations also supported these findings by comparing AI-based robotic systems with typical automation operations. Indices such as the efficiency of the production line, utilization of energy, and occurrence of faults indicated a 25% improvement in throughput and a 20% reduction in wastage of resources, demonstrating the quantitative benefits of embracing AI in manufacturing processes.

  Quantitative data of AI-based automation systems in manufacturing environments were demonstrated to yield a 20% reduction in energy consumption and a 15% enhancement in overall yield efficiency, illustrating the measurable benefits of AI-based optimization.

  Other learnings are from the food processing industry, where AI-enabled robotics improved sorting efficiency by 25%, thereby reducing food waste and operational costs. In the pharma sector, AI-enabled automation has accelerated the drug manufacturing process, with a reduction in batch production time by 15% while maintaining high regulatory compliance.

• 4.

  Development of intelligent infrastructure systems:

  - Strengths: Artificial intelligence applications in civil engineering are significantly influencing urban infrastructure through optimized traffic management, improved project efficiency, and better structural health monitoring. Such technologies result in safer, more efficient, and sustainable cities.

  An example in the hospitality industry shows that hotels’ AI-based building management systems have achieved up to 18% energy savings by optimizing Heating, Ventilation, and Air Conditioning (HVAC) and lighting systems in real time. Additionally, in logistics hubs, AI-based traffic optimization has reduced vehicle idling times by 20%, lowering emissions and improving throughput efficiency.

  - Weaknesses: It can be challenging to integrate AI into existing infrastructures due to legacy systems and the need for high levels of upgrades. Additionally, managing the vast amounts of data generated by smart infrastructure systems requires highly efficient data storage and processing solutions, which may be expensive and complex to develop.

  - Application of artificial intelligence in smart infrastructure has also been further substantiated by simulations that have demonstrated traffic flow improvement and project management effectiveness, along with statistical data demonstrating decreases in construction cost and time.

  Kim et al. found in a large-scale experiment that the utilization of AI algorithms for traffic flow control in smart cities reduced congestion time by 30% and improved road safety by 25% owing to their ability to streamline traffic patterns in real time.

  Further statistical comparison between AI-tuned traffic networks and traditional models revealed a 40% reduction in average traffic delays and a 15% boost in urban fuel efficiency in tested cities. These statistical metrics illustrate the transformative impact of AI on city planning.

  Furthermore, statistical analysis of AI-enhanced construction management systems indicated a 12% reduction in project postponement and a 20% reduction in material cost, exemplifying the capability of AI to promote construction efficiency.

• 5.

  Applications of environmental engineering:

  - Strengths: In the field of environmental engineering, AI is crucial for use in ecosystem monitoring, pollutant level predicting, and the management of natural resources. They are of pivotal relevance in world-scale problem-solving activities like climate change and resource depletion. In farming, AI-based irrigation systems have been shown to enhance water use efficiency by 30% through dynamic modulation of water supply in accordance with real-time crop and climatic conditions. Moreover, AI models employed for waste management have optimized recycling operations to achieve a 25% increase in material recovery in municipal waste treatment facilities.

  - Weaknesses: The setting is very complicated, and data quality is questionable; it can affect AI modeling accuracy and reliability. There are also ethical concerns, especially how AI can help manage natural resources without prejudice to fairness and equity in resource allocation.

  - Different studies have shown that statistical models and simulations supported by artificial intelligence are now essential tools in environmental engineering. In particular, they allow for improved accuracy in pollution prediction and optimization of resource management strategy.

  A recent study by Li et al. indicated that the use of AI-based models in air quality prediction enhanced the precision of pollution forecasts by 35%, allowing the authorities to initiate early measures in reducing emissions.

  To confirm these findings, simulation-based studies were carried out on real-time sensor data for predicting pollution levels under varying environmental conditions. AI algorithms outperformed conventional methods by achieving a 20% increase in accuracy and reducing false positive alarms by 15%, thereby further confirming their credibility in environmental monitoring.

  Further, quantitative analysis revealed that artificial intelligence-driven water resource management systems are able to enhance the efficiency of allocation by 18% amidst drought through the utilization of historical weather conditions and usage.

• 6.

  Technological innovations in electrical and energy engineering:

  - Strengths: AI is being used in electrical engineering to advance areas like renewable energy management and smart grids. These technologies are rendering power distribution networks efficient and environmentally friendly.

  They optimize the consumption of energy, reduce wastage, and enable the incorporation of renewable sources of energy.

  Detailed case studies demonstrate that artificial intelligence-based demand response systems in commercial buildings have cut peak energy usage by 15% and, thereby, enhanced grid stability. Likewise, in wind power plants, artificial intelligence has enhanced turbine efficiency by 10% through real-time adjustment of blade angles according to predictive weather forecasting.

  - Weaknesses: The integration with current power systems is challenging due to the fact that the older power systems were not originally designed to be integrated with such advanced technology. Additionally, data security and privacy issues are serious concerns in these applications because the presence of critical infrastructure is involved.

  - It has been empirically confirmed through simulations and statistical analysis that AI in electrical and energy engineering significantly improves the efficiency of energy distribution and facilitates the integration of renewable energy resources.

  Ricciardi et al. illustrated that the integration of AI into smart electrical grids reduced energy loss by 10% and added capacity to the grids for renewable sources like solar and wind, thereby enhancing the stability of the electrical system as a whole.

  Quantitative benchmarking of AI-based smart grids with traditional grid management systems revealed 30% improvement in the efficiency of load balancing and a 25% reduction in the outage duration. Such parameters demonstrate the potential of AI in revolutionizing energy distribution networks as well as its ability to embrace the integration of renewable energy.

  Statistical simulation of AI-driven renewable energy forecasting models reflected a mean 15% better accuracy in forecasting compared to traditional techniques, which resulted in a drastic reduction in the utilization of backup energy resources.

• 7.

  Advancements in aerospace and space exploration:

  - Strengths: AI deployment in aerospace enhances safety and operational effectiveness by way of autonomous systems and predictive maintenance. AI-powered robots and rovers are key to space exploration, offering autonomous navigation and data analysis on other planets.

  - Weaknesses: The strength and dependability of AI systems in extreme conditions, such as in space or at high altitudes, must be experimented with. Moreover, the extremely high research and financial inputs that are required to develop these systems form a barrier to mass adaptation.

  - AI has been proven to improve safety and efficiency in aerospace, particularly in autonomous flight systems and predictive maintenance, which reduces operational risks.

  Smith et al. found through a study that AI algorithms used for predictive maintenance of aerospace components reduced the risk of in-flight failure by 40%. This has significantly improved safety and operational efficiency for airlines.

  Additionally, the application of AI by NASA on autonomous rovers has provided 25% improved mission effectiveness with respect to optimized route mapping and real-time obstruction avoidance.

  In the defense sector, AI-powered autonomous drones have successfully carried out surveillance missions with 30% greater accuracy in target identification compared to traditional systems.

  Additionally, in the commercial aviation industry, case studies show that predictive maintenance systems have extended the life of major parts by 15%, reducing the number of replacements.

3.3. Multidisciplinary Applications of Artificial Intelligence in Engineering

AI technologies have made substantial contributions to one of the oldest and most general engineering disciplines: mechanical engineering. In mechanical engineering optimization, AI-based design generates more optimal and innovative designs by using algorithms. Generative design, an AI, uses machine learning to generate multiple alternatives for a given design and then make sense quickly in an attempt to get the best form. But AI also decreases the cost and time of mechanical part design and enhances their performance too [4]. Autodesk, which has had this method thrust upon them from the start, allows us to produce forms that, although possibly far superior to anything produced by human ingenuity, are frequently impossible to achieve using manual techniques.

Another important application is predictive maintenance, in which AI predicts equipment failures before they happen, reducing downtime and maintenance costs. Machine learning algorithms analyze historical data to anticipate potential issues, allowing proactive maintenance schedules to be implemented.

For example, neural networks monitor the fitness of machines in real-time and warn of any likely malfunction ahead of time [5]. Technologies like these are now widely used in industries such as aerospace and automotive for continuous and glitch-free operations.

Apart from that, AI technologies such as robotics and automation have transformed manufacturing operations. AI-driven sophisticated robotics can perform very complex operations with precision and consistency. Machine learning algorithms fine-tune production lines to provide better efficiency and minimal waste. A good example is AI applied to additive manufacturing (3D printing), where AI fine-tunes printing parameters to provide better quality and strength of printed parts [6,23].

For example, in medicine, AI-enhanced 3D printing is utilized to fabricate personalized prosthetics and implants based on the individual patient’s requirements.

In the coming years, AI will have a vital role in accelerating the identification of new materials with desired properties. Machine learning techniques have been developed to predict material properties from their atomic structure, reducing the need for time-consuming and expensive experiments. For example, Jha et al. [24] used machine learning to discover new battery materials for high-capacity batteries and significantly improved the research and development cycle. The ability to make rapid predictions and design new materials is particularly useful in fields like energy and manufacturing, where rapid innovation can be the basis of a significant competitive advantage.

The development of AI in civil engineering leads to intelligent infrastructure through digital technology being infused into physical infrastructure. Through AI power, intelligent sensors are applied to monitor data analytics that help optimize infrastructure performance and maintenance. For instance, AI algorithms regulate traffic flow in smart cities to reduce congestion while ensuring safety [7]. AI’s ability to process enormous amounts of data and provide real-time insights drives the urban planning and infrastructure management revolution to make cities sustainable and habitable.

Figure 4 is a heatmap representation of the impact of AI on traffic optimization, where AI-driven algorithms dynamically control traffic and reduce congestion in urban scenarios.

AI also enhances construction management by better project planning, scheduling, and resource allocation. Machine learning algorithms predict project outcomes based on historical data, leading to better decision-making [25,26].

Drones and robotically driven equipment powered by AI are now an absolute necessity where site inspection and monitoring are concerned, with more precision and enhanced safety [8]. They are transforming the construction industry through fewer chances of human error, increased efficiency, and raising the safety bar higher.

In structural health monitoring, AI is vital in determining the status of infrastructure and forecasting expected issues. With information from sensors incorporated in structures, AI software can detect anomalies and determine overall integrity.

This vision not only works to ensure the safety of infrastructure work but also increases its life-cycle [12]. See how AI is used in monitoring bridges and tunnels to receive proactive alerts as well as an accurate replica of maintenance recommendations in case a fault is detected.

The application of artificial intelligence (AI) technologies has already enabled much in the field of electrical engineering. And smart grids as AI-driven electricity distribution systems: more efficient, more secure, green [27].

AI optimizes electricity flow, control, and regulation to manage demand and efficiently incorporate renewable energies like solar or wind power into grids. The result, fewer energy losses and thus more stable grids [15]. AI is also crucial to coordinate renewable energy that can forecast how much clean energy it will produce by pitting machine learning against weather prediction, which is capable of maximizing the use of renewables and minimizing the use of fossil fuels [16]. For instance, Google’s DeepMind has documented working with energy companies to predict wind farm output lowers the cost of transmitting renewable energy and increases its value, in the process increasing reliability.

More AI is employed in renewable energy, like solar and wind. Artificial Intelligence learns how weather conditions impact the production and utilization of energy to utilize renewable energies more effectively without fossil fuels. This indirectly supports the prevention of failure for sun, water, and wind machinery [16]. Renewable energy units thus gradually become more reliable and cost-effective while facilitating the transition of the world towards green power sources.

In the domain of electronic design, AI has a gigantic influence via Electronic Design Automation (EDA)—where machines can learn to forecast cycle times and component characteristics from vast datasets comprising thousands or millions of parts.

EDA tools using AI automatically design electronic systems and circuits, optimizing the performance and layout of components, thus saving time and cost [28,29].

These AI technologies also make it possible to have more advanced and more effective electronic systems [18]. In the semiconductor industry, for instance, AI is used to create intricate chip designs that push the limits of computing performance and efficiency.

Aerospace engineering, the field in which aircraft and spacecraft are designed and manufactured, has been greatly boosted by AI. AI-powered autonomous flight systems are used to heighten the safety and efficiency of aircraft flight operations. Machine learning algorithms analyze high volumes of flight data to identify optimized routes to take, conserve fuel, and land safely. Autonomous drones propelled by AI are also widely utilized to carry out surveillance and package delivery [19].

In aviation, predictive maintenance is another area where AI shines. Tracking the health of aircraft components, AI can predict potential failures. AI can detect signs of wear and tear in aircraft components by analyzing data from sensors within the aircraft and carry out maintenance before failure [21].

AI is revolutionizing transportation engineering by optimizing traffic management, vehicle safety, and making autonomous driving possible. AI-driven traffic management systems utilize live data from cameras and sensors to control traffic movement and reduce congestion [30]. When it comes to vehicle safety, AI-driven advanced driver-assistance systems Advanced Driver-Assistance Systems (ADAS) prevent collisions and monitor drivers, significantly eliminating the possibility of accidents [31].

The development of self-driving cars also depends heavily on AI technologies like computer vision and deep learning that enable cars to navigate through challenging environments and decide in a fraction of a second [32]. Tesla and Waymo are among the leading firms working on creating AI-driven self-driving cars that will transform transportation in the future. In Table 10, all the facts discussed in this section are summed up.

In aerospace engineering, predictive maintenance applies AI to track aircraft component health closely and forecast probable issues. AI programs execute analysis on data input from sensors embedded in the plane to pick up minute indications of fatigue that allow for preventive action and to eradicate the possibility of in-flight failures [21].

AI plays a growing role in space exploration, as AI-driven robots and rovers take on significant assignments like planetary reconnaissance, sampling, and data analysis. Machine learning algorithms filter vast amounts of data from space missions, delivering valuable information and improving mission outcomes [22]. NASA’s Mars rovers, for example, utilize AI to navigate autonomously across the Martian terrain, finding areas of interest to further explore and establish their most effective routes.

Integration with advanced technologies, such as the Internet of Things (IoT) and blockchain, is transforming the process of engineering. The data from IoT can be directly processed through AI algorithms to make decisions with precise judgments of the circumstances. In intelligent grids, it utilizes IoT data to optimize energy supply dynamically in time to improve operation to make it more efficient and sustainable. While this, blockchain provides incredibly robust security against transparency to the operations of AI, which is critical for use cases involving sensitive data or affecting infrastructure at the core. This power couple is making engineering processes faster, safer, and more innovative while combining quicker, safer [33]. For instance, a blend of blockchain and AI is piloted in supply chain management to increase transparency, traceability and efficiency.

In collaborative systems, newer paradigms in AI, like federated learning and reinforcement learning, offer novel solutions to counter problems arising in distributed data environments and complex decision-making. Federated learning enables secure and cooperative data analysis across several stakeholders, and reinforcement learning enables adaptive decision-making in dynamic and interconnected systems, paving the way for more robust and efficient engineering solutions.

It maximizes processes, enhances safety, and improves material synthesis; AI is transforming chemical engineering. The results of reactions can be forecasted by machine learning algorithms [34], which could help improve the efficiency and sustainability of chemical processes; Material synthesis AI helps accelerate the discovery of new catalysts and materials with specific properties by using sophisticated pattern recognition on extremely large amounts of experimental data. Apart from this, AI enhances process safety by monitoring the chemical plants in real-time and providing advanced warnings of potential hazards to allow companies to take preventive measures in advance [35,36]. All the above information presented in this section is presented in Table 11.

These new computer programs are able to view with a virtual eye as chemical reactions happen and make changes along the way, while providing warnings if conditions seem dangerous.

Biomedical engineering experiences a revolution phase with the abilities and applications that AI has, replacing traditional medical treatments with solution mechanisms for diagnosis, treatment care recordings for patients. For medical imaging, AI has the key advantage of integration with Magnetic Resonance Imaging (MRI), CT, and X-rays with total precision, thus able to catch in a moment any anomaly that could go undetected by human eyes. For example, Litjens et al. In another study [37], researchers achieved 15% higher accuracy in breast cancer detection with the assistance of a convolutional neural network (CNN) method.

In prosthetics, AI is also making new ground. Custom prosthetic limbs that are lighter in weight, yet don’t compromise in strength, can even be crafted by implementing machine learning models on one’s patient data to analyze it. With this prosthetic functional upgrade, their findings are aligned with Hensman et al., as evidenced that the quality of an individual should be improved among such patients [38].

Furthermore, AI has been instrumental in developing the science of personalized medicine. Through examining complex medical and genetic information, AI can help develop tailored treatment strategies that lead to much better patient outcomes.

Google DeepMind’s artificial intelligence system for medical imaging is an excellent example of this revolution. Using convolutional neural networks (CNNs), the system has increased early detection rates for retinal disease by 15%, outperforming traditional diagnostic methods. This demonstrates how diagnostic accuracy is enhanced and the likelihood of missed anomalies lowered by AI. All the information explained in this section is presented in Table 12.

For example, Kourou et al. [39] used predictive models in creating personalized treatment plans for cancer, which led to remarkably enhanced patient outcomes.

The potential for AI in personalized medicine is extremely bright. With the combination of AI, genomics, and wearable devices, treatment protocols can be very personalized based on a person’s own genes and daily patterns.

AI applications can also analyze real-time data from wearable sensors to monitor a patient’s health and make personalized therapeutic recommendations. This not only enhances the effectiveness of treatments but also enhances patient adherence to medical prescriptions, yielding better overall clinical outcomes [40].

3.4. Next-Generation AI Applications in Engineering

Though the paper covers a broad spectrum of established fields, it must also be mentioned that there are novel emerging applications of AI in new areas like biomedical engineering and materials science. In biomedical engineering, AI is transforming medical imaging, prosthetic design, and personalized medicine. AI algorithms can potentially analyze complex biological data to create personalized treatment plans, enhancing patient outcomes [40]. Moreover, frontier applications such as genomics and materials science AI are representative of the potential for paradigm-shifting innovation. Representative examples, performance metrics, and computational requirements for these applications are summarized in Table 13.

AI is speeding up the discovery of novel materials in material science by predicting how they would respond based on their atomic configurations. This significantly speeds up innovation [33]. AI allows scientists to discover new materials much faster than traditional means. For example, Jha et al. [24] used machine learning to discover battery materials with high capacities, significantly reducing research and development time. AI is also critical in predicting material properties from their atomic structure, an activity that normally requires long and complex experiments [41].

In addition, Cloud-based AI platforms have also become central to expediting material discovery by allowing researchers to scale computational simulations over distributed infrastructures. Moreover, edge computing enables real-time monitoring and optimization of material manufacturing processes at reduced costs and increased efficiency. These developments emphasize the role of scalable AI systems in resolving both computational and operational issues in materials science.

Xie and Grossman [42] were able to harness this potential with a model that accurately predicts thermal conductivity in materials, demonstrating the ease with which AI can facilitate discovery. AI is also making manufacturing processes more efficient and cost-effective. Zhang et al. [43] used machine learning to determine optimal 3D printing parameters for composite materials, leading to parts of greater quality and strength. This maximization not only improves the manufacturing process but also contributes to the production of improved materials and products.

When such new uses gain traction, they create opportunities for innovative engineering solutions that span both conventional problem-solving and frontier possibilities. All the content covered in this section is summarized in Table 14.

3.5. Research Gaps and Future Pathways

As AI continues to evolve, several new challenges and directions for the future have emerged from the discourse of contemporary approaches. These challenges and new directions will be key to driving further progress in AI in engineering.

• 1.

  New challenges:

- Data Privacy and Security: As AI-driven systems handle sensitive and large amounts of datasets, data privacy and security have become a top priority. There is a growing need for advanced encryption methods, secure data storage warehouses, and policies that comply with international data protection laws.

- Scalability and Integration: Scaling AI solutions to be effective in real, large-scale engineering systems is challenging. Integrating AI with existing legacy systems requires a comprehensive infrastructure upgrade and can face resistance from stakeholders used to traditional methods. This project underscores the fact that the absence of a universal assessment framework further complicates the implementation of AI since engineers cannot compare and verify AI solutions in evolving, multi-disciplinary environments.

In distributed data systems, federated learning is a promising solution that emerged to advance collaborative model training and data privacy protection. Reinforcement learning also has the potential to enhance system flexibility and decision-making in complex, dynamic scenarios and enable the operation of AI systems optimally on interconnected infrastructures. These paradigms address basic challenges while opening new space for scalable and responsible AI incorporation.

- Bias and Fairness in AI: AI systems are prone to biases in training data, which may result in unfair or suboptimal outcomes in critical engineering applications. Data diversity must be ensured, and algorithms must be designed to reduce bias for the ethical use of AI technology.

• 2.

  Pathways for future research:

- Integration with Emerging Technologies: Combining AI with emerging technologies like the Internet of Things (IoT) and blockchain opens up new possibilities. For example, IoT can offer real-time data that improves the accuracy of AI-driven decisions in smart infrastructure, while blockchain can offer security and integrity to these processes.

- Advancements in AI-Driven Design and Optimization: AI-based design methodologies need to be minimized in terms of computational needs so that they can be applied across industries on a larger scale. Additionally, the development of new algorithms capable of handling uncertainty and variability in design parameters will be pivotal in driving further innovation.

- Enhanced Human-AI Collaboration: Creating guidelines for enabling more effective collaboration among human engineers and AI systems can lead to more innovative solutions. This would involve AI systems offering real-time critiques, suggesting design alternatives, or assisting in making decisions, hence enhancing engineering creativity and problem-solving skills.

These issues and directions not only highlight the current limitations of AI in engineering but also predict the upcoming developments needed to tap its maximum potential.

3.6. Support with Statistical Simulation and Analysis

The conclusions and assertions made in this survey are substantiated by statistical modeling and simulation where necessary. For instance, the effectiveness of AI-driven predictive maintenance has been illustrated through analysis of machine failure rate data for industry, showing a precipitous drop in downtime and maintenance cost. Similarly, simulations of AI-driven design optimization against conventional methodology have yielded efficiency as well as design quality gains.

To further confirm the discussion, statistical testing has been used to evaluate the performance of various AI techniques in various engineering tasks. Such analyses provide strong quantitative backing for the qualitative judgments made during the survey, such that conclusions derived are valid and reliable. For the future, this kind of research must continue to refine statistical and simulation-based methods to support and improve AI applications in engineering so that they can become useful and functional in actual environments.

Besides, statistical studies of federated learning approaches have demonstrated their efficacy in maintaining model performance while maintaining data privacy, particularly in distributed healthcare and industrial environments. Likewise, testing on reinforcement learning in manufacturing has registered notable drops in energy usage and operational waste, indicating the flexibility of the techniques in dynamic systems.

4.1. Challenges in Evaluating AI Performance Across Engineering Domains

A commonly neglected significant problem with the application of artificial intelligence systems in engineering is the potential for overestimation of performance when testing is restricted to a particular domain. The metrics applied in niche domains, for instance, the accuracy of predictive maintenance models for manufacturing industry applications, may not be readily comparable with those applied to electrical systems or civil infrastructure.

This lack of consistency in assessment can create unrealistic expectations and hinder the transferability of AI software between industries. Therefore, it is necessary to create standard protocols for performance assessment that can be used across engineering disciplines.

4.2. Ethical and Social Implications

There will, nevertheless, be intense social and ethical questions that arise with the application of AI in engineering. Some of these are issues that job automation will affect work and the responsibility of AI-driven decisions, whether or not this type of technology is ethically compliant. Alternatively, automation of activities traditionally performed by humans may lead to job loss, which consequently needs reskilling and workforce transitioning strategies [50].

Moreover, decision-making processes of AI systems should be transparent and make the behavior understandable to ensure faith in engineers and stakeholders [51].

Yet another ethical concern is the potential bias inherent in AI algorithms that may lead to discrimination or discriminatory outcomes. To negate this, training data must be formulated with a conscientious regard for diversity [49] and anti-bias techniques in all development stages.

All the information provided in this section is presented in Table 15.

5.1. Standardizing Evaluation Procedures for AI Implementation in Industries

One key success factor in scaling the impact of AI across diverse engineering disciplines will be the development of standard test protocols. The protocols must define common performance metrics to enable comparison of AI models across application contexts, transfer of solutions, and mitigation of performance overestimation risk. The adoption of common standards can also shrink validation cycles, lower development costs, and simplify decision-making for engineers working in multidisciplinary environments.