Modern Revolutions in Science, Engineering and Technological Researches with the Developments of Artificial Intelligence (AI) Technology, Tools and Applications: A Review

Aluvihara S, Faqiri AU, Pestano-Gupta F, Hekmat AW, Hekmat AW and Karimkhani M

Published on: 2026-09-26

Abstract

The advent and rapid evolution of Artificial Intelligence (AI) have instigated a profound transformation across the landscapes of science, engineering, and technological research. Far from being merely a sophisticated tool, AI, encompassing machine learning, deep learning, natural language processing, and computer vision, has emerged as a co-revolutionary force, fundamentally altering methodologies, accelerating discovery, and enabling unprecedented capabilities. This review article comprehensively examines how AI technologies, tools, and applications are catalyzing modern revolutions in diverse scientific and engineering disciplines. We explore AI's impact on hypothesis generation, data analysis, experimental design, and the synthesis of new knowledge, highlighting its role in domains such as materials science, healthcare, advanced manufacturing, environmental modeling, and fundamental physics. Furthermore, we delve into the specific AI paradigms and computational tools that are driving these advancements, from generative adversarial networks and transformers to robotics and digital twins. Finally, this article addresses the inherent challenges, ethical considerations, and the future trajectories of AI in shaping the trajectory of human innovation, emphasizing the necessity for responsible development and interdisciplinary collaboration to harness AI's full potential for a prosperous future.

Keywords

Artificial Intelligence; Machine Learning; Deep Learning; Scientific Discovery; Engineering Research; Technological Innovation; Healthcare; Materials Science; Robotics; Data Science

Introduction

The 21st century has seen a remarkable merging of several key trends: a huge growth in available data, big jumps in the power of computers, and new ideas in how algorithms work. These trends have led to something new: the growth of Artificial Intelligence (AI). What started as a small area of research focused on making machines thinks like people has now become a widespread part of our lives. AI is changing how we do science, design things, and create new technologies [1]. The way we usually do science, by watching, guessing, testing, and thinking, is changing. AI is becoming a helpful partner in each of these steps.

Today's changes in science, engineering, and technology are happening because of both human thinking and help from AI. AI can handle large amounts of data, spot hidden patterns, guess what might happen, and come up with new answers faster than people can [2]. AI is not just making things work better; it is allowing us to discover and create in whole new ways.

For example, AI is helping us design new materials, give people medicine that is tailored to them, improve complicated engineering systems, and understand mysteries of the universe [3].

This article will look at how AI is changing science, engineering, and technology. It will explain the main AI technologies that are behind these changes. It will also show specific examples of how AI is used in different fields. It will talk about the tools and systems that make it easier to use AI. Finally, it will discuss the problems, ethical questions, and possibilities for the future of AI. The goal is to show how AI is not just a tool but a basic force that is changing how the world innovates. Additionally, figure 1 illustrates the integration of AI into the scientific research process, from the acquisition of data to knowledge generation.

Figure 1.A Conceptual framework of an AI-driven scientific research cycle.

The figure 1 shows the progression from data integration to knowledge discovery, with iterative feedback and human-AI collaboration enhancing decision-making and model refinement.

The AI Change, Basic Technologies in Research

The term Artificial Intelligence covers many different ideas, methods, and technologies. These are all aimed at making machines do things that usually need human intelligence. The current AI boom is mostly because of progress in certain areas [4].

Machine Learning (ML)

ML is a key idea in AI. It involves algorithms that learn from data without specific instructions. It means creating models that can use what they have learned to make guesses or decisions about new data. The main ML methods include the followings.

Supervised Learning

This involves learning from data that has labels (input-output pairs) to predict what will happen in the future. It is used a lot in areas like finding diseases from medical images or guessing what properties a material will have [5].

Unsupervised Learning

This involves finding patterns or structures in data that does not have labels. It is used for things like grouping patients into subgroups or simplifying complicated scientific data [6].

Reinforcement Learning (RL)

This involves training machines to make decisions in an environment to get the best reward. RL has been very successful in improving complicated systems, robotics, and scientific experiments [7].

Deep Learning (DL)

DL is a special type of ML that uses artificial neural networks with many layers. This is why it is called deep. DL can automatically find important features in raw data. It has led to many recent breakthroughs in AI because of this ability.

Convolutional Neural Networks (CNNs)

CNNs are mostly used for looking at images and videos. They have changed computer vision tasks like recognizing objects, classifying images, and analyzing medical images [8].

Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM)

RNNs are designed to work with sequential data. This makes them good for things like understanding language, predicting time series, and analyzing genetic sequences [9].

Transformers

Transformers are a newer type of architecture that is very good at language tasks. They are behind large language models (LLMs) like GPT-3/4 and BERT. These models can understand and create text that is very similar to human writing [10].

Generative Adversarial Networks (GANs)

GANs have two parts: a generator and a discriminator network. They are good at making synthetic data, such as images, and material structures. This data often looks very real [11].

Natural Language Processing (NLP)

NLP is about making computers understand, interpret, and create human language. It is used for things like analyzing scientific papers, coming up with new ideas, searching for information, and writing reports automatically [12].

Computer Vision (CV)

CV gives computers the ability to see and understand information from the world. In research, CV is important for things like automated microscopy, analyzing remote sensing data, finding defects in manufacturing, and studying behavior in biology [13].

These basic AI technologies are often used together. They are the foundation for the advanced tools and uses that are changing how research is done in many fields.

How AI Is Changing Science and Engineering

AI is being used in many different scientific and engineering fields. It is speeding up discoveries and making it possible to do things that were not possible before.

Figure 2. The schematic representation of artificial intelligence.

Figure 2 interprets the artificial intelligence as a central driver of innovation across multiple scientific domains, including engineering, basic sciences, environmental science, healthcare, and materials science.

Materials Science and Chemistry, Speeding Up Discovery and Design

AI is changing how we find and design new materials and chemical compounds. This process used to be slow and involve a lot of trial and error.

Accelerated Materials Discovery

ML models can guess what properties a material will have based on its atomic structure. For example, they can predict conductivity, strength, and thermal stability. This helps narrow down the search for new materials [14]. This method, called materials informatics, uses large databases of known materials to train AI models. These models can then design materials with specific properties [15].

Drug Discovery and Development

AI is changing every step of the drug development process. It can find potential drug candidates by screening millions of molecules using deep learning. It can also predict how well a drug will work, how toxic it is, and what side effects it might have. This greatly reduces the time and money it takes to bring new drugs to market (Fleming, 2018). AlphaFold, a deep learning model from DeepMind, can accurately predict protein structures. This is a big breakthrough that speeds up our understanding of biological processes and how to find drug targets [16].

Catalyst Design and Improvement

AI can predict how well catalysts will perform. Catalysts are used to speed up chemical reactions. AI can optimize reaction conditions and even design new catalytic materials that are more efficient and selective.

Automated Synthesis and Characterization

Robotics and AI are being used to create autonomous laboratories. In these labs, machines can carefully design, carry out, and analyze experiments. This reduces human error and increases the amount of work that can be done [18].

Healthcare, Medicine, Precision, Prediction, and Personalization

AI is leading a medical revolution. It is allowing for more accurate diagnoses, personalized treatments, and better healthcare.

Diagnostic Imaging and Pathology

CNNs are very good at looking at medical images, such as X-rays, MRIs, CT scans, and pathology slides. They can spot small problems that indicate diseases like cancer, diabetic retinopathy, or pneumonia. They often do this as well as or better than human experts [19].

Personalized Medicine

AI models can combine information like genomic data, electronic health records (EHRs), lifestyle choices, and environmental factors. This allows them to predict how people will respond to treatments, find the best drug dosages, and forecast how diseases will progress. This is leading to truly personalized healthcare [20].

Epidemiology and Public Health

AI models can predict disease outbreaks, track how they spread, and assess how well public health actions are working. This was seen during the COVID-19 pandemic [21].

Drug Repurposing and Clinical Trials

ML algorithms can find existing drugs that might work against new diseases. This shortens the drug discovery process. They can also improve how clinical trials are designed, how patients are recruited, and how outcomes are predicted [22].

Robotics in Surgery and Care

AI-powered surgical robots can improve accuracy, reduce how invasive surgeries are, and help surgeons. Companion robots can care for the elderly and disabled, improving their quality of life [23].

Engineering Sciences: Design, Improvement, and Smart Systems

AI is changing engineering from the beginning to the end of the process, including design, manufacturing, and operations.

Generative Design and Improvement

AI algorithms can automatically create many different designs for mechanical parts, structures, or circuits. They can optimize for things like weight, strength, cost, or energy use. This allows engineers to explore designs that they might not have thought of on their own [24]. Topology improvement, which is further improved by AI, creates complex, high-performing structures.

Advanced Manufacturing and Industry 4.0

AI is used in smart factories for things like predictive maintenance (finding when equipment is likely to fail), quality control (automated visual inspection), robotic automation, and supply chain improvement. This leads to better efficiency, less waste, and improved safety [25].

Robotics and Autonomous Systems

AI is important for developing autonomous vehicles, drones for inspection and delivery, and advanced robotic systems for dangerous environments. This improves safety and accessibility [26].

Smart Infrastructure and Energy Systems

AI improves energy grids, manages smart city infrastructure (traffic, waste, public services), and increases the efficiency of renewable energy systems by predicting changes in demand and supply [27].

Aerospace and Defense

AI helps with aerospace design (like improving aerodynamics), flight control, satellite image analysis, and intelligent defense systems. This increases performance and operational safety [28].

Environmental Science and Earth Sciences in Watching, Predicting, and Reducing Problems

AI gives us tools to understand, watch, and reduce environmental problems and climate change.

Climate Modeling and Prediction

AI improves climate models by processing large amounts of satellite images, sensor data, and past climate records. This improves the accuracy of weather forecasts, climate change predictions, and disaster predictions.

Biodiversity Watching and Conservation

CV and ML algorithms can identify species from camera trap images, watch deforestation, track animal movements, and analyze environmental DNA. This provides important data for conservation efforts [29].

Pollution Watching and Resource Management

AI analyzes air and water quality data to find pollution sources, predict when contamination might happen, and improve how we use resources like water, agriculture, and waste management [30].

Renewable Energy Improvement

AI makes solar, wind, and geothermal power generation more efficient by predicting energy patterns and improving grid integration [31].

Basic Sciences, Finding the Universe's Secrets

AI is being used more and more in basic scientific research, from the smallest particles to the largest scales of the universe.

Physics and Astronomy

AI helps analyze large datasets from particle accelerators to spot new particles or events, detect gravitational waves, classify galaxies, and find exoplanets from telescope data [32]. ML algorithms are used to speed up complicated simulations in fluid dynamics, plasma physics, and quantum mechanics [33].

Biology (beyond medicine)

AI is very important in genomics and proteomics for analyzing sequences, predicting gene function, mapping protein-protein interactions, and understanding complex biological networks. It speeds up the analysis of single-cell RNA sequencing data, gives insights into evolutionary biology, and helps design synthetic biological pathways [34].

Chemistry (beyond materials)

AI helps predict chemical reactions, improve synthesis pathways, and understand complex molecular interactions. This These technologies open new ways for basic chemical research.

AI Technologies, Tools, and Applications

Besides its impact on different fields, specific AI technologies and tools are changing the everyday work and methods of scientific and engineering research.

Figure 3. Some key challenges associated with the application of AI technology.

Figure 3 interprets the key challenges associated with the applications of AI technology in scientific research, including data constraints, ethical considerations, lack of model interpretability, computational limitations, restricted generalizability, and reproducibility issues.

Automated Scientific Discovery Platforms

AI is leading to the creation of research systems that can generate ideas, design experiments, carry them out with robots, analyze results, and even change ideas in cycles.

Robot Scientists

Systems like Adam and Eve (for microbiology) combine AI with robotic automation to do many experiments and test ideas without people. This speeds up the pace of discovery [35].

AI for Idea Generation

NLP models analyze many scientific papers to find hidden connections, predict new relationships between things (like genes and diseases or materials properties), and suggest new research directions [36].

Generative AI for Synthesis and Design

Generative AI models, especially GANs and Diffusion Models, are becoming powerful tools for creating new things that meet specific design needs.

Molecule and Material Generation

AI can create new molecular structures with specific properties, reducing the need for many experiments. This is important in drug design (de novo drug design) and advanced materials [37].

Synthetic Data Generation

For areas with limited or sensitive data, generative models can create realistic synthetic datasets for training other AI models, protecting privacy and increasing training possibilities.

Architectural and Engineering Design

AI generates new designs for buildings, bridges, and mechanical parts, exploring ideas that human designers might not think of.

Natural Language Processing (NLP) for Knowledge Finding and Combination

The large amount of scientific papers makes it impossible for anyone to keep up. NLP is changing how researchers work with knowledge.

Automated Literature Review

NLP tools can summarize papers, find key information, spot trends, and combine information from different sources to create reviews or answer scientific questions [38].

Semantic Search and Knowledge Graphs

Instead of just matching keywords, NLP-powered semantic search engines understand the meaning behind questions to find relevant scientific information. Knowledge graphs, created through NLP, map relationships between scientific things (genes, proteins, diseases, compounds), allow for advanced knowledge discovery [39].

Scientific Writing Help

AI writing assistants can help researchers write papers, improve grammar and style, and even create initial text based on research data.

Robotics and Automation for Faster Experiments

AI and advanced robotics are working together to automate and speed up experimental work.

High-Throughput Experimentation

Robotic systems, guided by AI, can do many experiments at the same time, testing different conditions for drug screening, materials synthesis, or chemical reactions [40].

Autonomous Laboratories

AI controls robotic arms, liquid handlers, and analytical instruments to do entire experimental tasks, from preparing samples to collecting data. This frees up human researchers to do more advanced intellectual work.

Remote and Dangerous Environments

AI-powered robots can work in places that are not accessible or safe for humans, such as deep-sea exploration, space missions, or nuclear facilities. They can collect important data and do complex tasks.

Digital Twins and Simulation

Digital twins are virtual copies of physical things, processes, or systems. Combined with AI, they allow for real-time monitoring, predictive maintenance, and improved performance.

Process Improvement

In manufacturing, a digital twin can simulate the entire production line. This allows AI to find problems, predict failures, and improve settings for better efficiency and quality [41].

Infrastructure Management                             

Digital twins of smart cities or energy grids allow AI to simulate different situations (like power outages or traffic jams) and test solutions virtually before they are implemented.

Predictive Analytics and Anomaly Detection

AI can spot patterns and anomalies in large datasets. This is important for making proactive decisions.

Predictive Maintenance

AI analyzes sensor data from machines to predict when parts are likely to fail. This allows for preventative maintenance and reduces downtime.

Cybersecurity in Research Infrastructure

AI detects unusual network behavior, spotting potential cyber threats and protecting sensitive research data.

Quality Control

AI-powered vision systems detect defects in manufactured goods faster and more accurately than human inspection.

Problems, Ethical Questions, and Responsible AI Development

Even though AI has a lot of potential, its widespread use in research comes with problems and ethical questions that need to be addressed for its development to be responsible and fair.

Data Dependence and Quality

AI models are only as good as the data they are trained on.

Data Availability and Access

Many scientific datasets are private, separated, or simply don't exist in enough quantity or quality. The need for large, diverse, and well-organized datasets is a problem.

Data Bias

If training data shows past biases (like in medical records or engineering designs), AI models will continue and even increase these biases. This can lead to unfair or inaccurate results, especially in sensitive areas like healthcare or law enforcement [42].

Data Privacy and Security

Using sensitive data (like patient health records or private research data) raises privacy concerns and requires strong cyber security to prevent breaches.

Interpretability and Explain ability (XAI)

Many powerful AI models, especially deep neural networks, work like black boxes. This makes it hard for people to understand how they reach their conclusions.

Lack of Trust and Use

In important uses like medical diagnosis or autonomous systems, lack of transparency can reduce trust among experts and the public, which prevents them from being used.

Difficulty in Fixing and Improving

Without understanding why an AI model makes a wrong prediction, it is hard to find errors or make targeted improvements.

Regulatory Problems

The black box nature can cause problems for following regulations, especially in fields that require accountability and transparency. The growing field of Explainable AI (XAI) aims to solve this by creating methods to make AI decisions easier to understand [43].

Ethical Effects and Societal Impact

The widespread use of AI raises deep ethical questions.

Automation and Job Loss

While AI creates new jobs, it also automates tasks, which can lead to job loss in different areas, including research support roles [44].

Misinformation and Harmful Use

Generative AI can be used to create fake scientific data or misinformation that looks real. This can hurt scientific integrity. It can also be misused in autonomous weapons systems or surveillance, which is a serious concern [45].

Accountability and Responsibility

When an AI system makes a big mistake, it is important to decide who is responsible: the developer, the user, or the AI itself. Creating clear legal and ethical rules is important.

Reproducibility Problem

While AI can help with reproducibility, complex AI models and their data processes can also make it harder to reproduce results in science if they are not well documented and transparent.

Resource Intensity and Environmental Impact

Training and using large AI models takes a lot of computer resources and energy, which increases carbon emissions. The environmental effect of AI is a growing concern that needs sustainable solutions [46].

Skill Gap and Interdisciplinary Work

To use AI fully, researchers need to learn new skills in data science, programming, and AI methods. It is important to connect domain experts and AI specialists through strong interdisciplinary work.

To address these problems, researchers, policymakers, industry leaders, and the public need to work together to make sure that AI development is guided by fairness, transparency, accountability, and environmental sustainability.

Future Trends, the Changing Human-AI Partnership

The use of AI in scientific and engineering research is still developing. Several new trends and future directions promise even bigger changes.

Towards Artificial General Intelligence (AGI) for Scientific Discovery

While current AI is good at specific tasks, the ultimate goal of AGI is AI that can understand, learn, and use intelligence in many different tasks, like humans do. An AGI could greatly change scientific discovery by creating complex theories, designing detailed experiments, and combining knowledge from different fields without human help [47]. While this is still a future goal, progress in large language models and multi-modal AI systems suggests that we are building the pieces for more general scientific thinking.

Combining AI with Quantum Computing

The combination of AI and quantum computing, called Quantum AI, is a promising area. Quantum computers could speed up AI training, allow for more complex simulations, and solve improvement problems that are too hard for classical computers [48]. AI could also improve quantum algorithms and design quantum hardware. This combination could lead to breakthroughs in drug discovery, materials science, and basic physics.

Neuro-Symbolic AI for Better Thinking

Combining the strengths of deep learning (pattern recognition) with symbolic AI (logical thinking and knowledge representation) is a good way to create AI systems that are stronger, clearer, and easier to understand. Neuro-symbolic AI could improve AI's ability to understand scientific ideas, create explainable ideas, and combine experimental data with theoretical knowledge [49].

AI for Science, and Science for AI

This relationship will grow stronger. AI will continue to speed up scientific discovery, but scientific ideas and biological knowledge will also inspire the next generation of AI designs and algorithms. For example, understanding how the brain works could lead to AI systems that are more efficient and adaptable.

Federated Learning and Privacy-Protecting AI

As data privacy becomes more important, privacy-protecting AI methods like federated learning will be necessary. This allows AI models to be trained on data from different sources (like hospitals or research institutions) without the data leaving its source. This protects privacy while maximizing the usefulness of data for research [50].

Making AI Tools More Accessible

Creating user-friendly AI platforms and libraries will further make AI available to more people. This will allow researchers 67 AI expertise to use these tools in their fields, encouraging widespread innovation [51-56]. The future of research is increasingly a partnership between human minds and advanced AI. This partnership will require continuous learning, ethical planning, and connections between different fields to fully use AI as a partner for human progress.

Conclusion

The development and widespread use of Artificial Intelligence technologies have started a new period of scientific, engineering, and technological changes. As discussed in this review, AI is not just a small change but a big force that is changing how research is done. From speeding up drug discovery and designing new materials with smart algorithms to changing healthcare diagnoses and improving complex engineering systems, AI has become an important partner in all parts of modern innovation. The main AI ideas machine learning, deep learning, natural language processing, and computer vision are behind a wide range of tools and uses. These include generative AI models that create new design possibilities, robotic systems that automate complex experiments, digital twins that provide real-time information, and predictive analytics that forecast important results. This partnership between AI and specific knowledge is breaking down old barriers, allowing researchers to solve problems of great complexity and size, leading to discoveries that were not possible before.

However, there are problems along the way. Issues with data quality and bias, the black box problem of understanding how AI makes decisions, ethical questions about automation and responsibility, and the environmental effect of large AI models need careful thought and proactive solutions. Responsible AI development, ethical rules, and strong regulations are important to make sure that these powerful technologies are used for the good of humanity.

Looking to the future, the combination of AI with new technologies like quantum computing, the search for more general intelligence, and the linking of human and AI cognitive strengths promise even bigger breakthroughs. The future of scientific and engineering research is closely connected to the development of AI, creating a relationship where human creativity and AI's analytical power work together to find new knowledge and innovation. This will help us solve some of the most important global problems. The change is happening now, and AI is at the center, guiding us towards a future of faster discovery and big technological progress.

References

  1. Jordan MI, Mitchell TM. Machine learning: Trends, perspectives, and prospects. Science. 2015; 349: 255-260.
  2. Zador AMA critique of pure learning and what artificial neural networks can tell us about animal brains. Nature Communications. 2019; 10: 3770.
  3. Wooley MA. The artificial intelligence revolution and its impact on medical devices. Moffitt Medical Journal. 2018; 5: 1-5.
  4. Goodfellow I, Bengio Y, Courville A. Deep Learning. MIT Press. 2016
  5. Alpaydin E. Introduction to Machine Learning. MIT Press. 2002.
  6. Hastie T, Tibshirani R, Friedman J. The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Springer. 2017
  7. Minh V, Kavukcuoglu K, Silver D, Rusu AA, Veness J, Bellemare M, et al. Human-level control through deep reinforcement learning. Nature. 2015; 518: 529-533.
  8. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015; 521: 436-444.
  9. Hochreiter S, Schmidhuber J. Long Short-Term Memory. Neural Computation. 1997; 9: 1735-1780.
  10. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Aidan N, et al. Attention Is All You Need. Advances in Neural Information Processing Systems. 2017; 30.
  11. Goodfellow I, Mirza M, Xu B, Warde-Farley D, Pouget-Abadie J, Ozair S, et al. Generative Adversarial Nets. Advances in Neural Information Processing Systems. 2014; 27.
  12. Young T, Hazarika D, Poria S, Cambria E. Chinese natural language processing: Recent advances and challenges. Future Generation Computer Systems. 2018; 89: 753-768.
  13. Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, et al. ImageNet Large Scale Visual Recognition Challenge. International Journal of Computer Vision. 2015; 11: 211-252.
  14. Raccuglia P, Elbert KC, Adler PDF, Falk C, Wenny MB, Mollo A, et al. Machine-learning-assisted discovery of new thermoelectric materials. Nature. 2016; 537: 636-640.
  15. Tshitoyan N, Dagdelen J, Weston L, Dunn A, Rong Z, Kononova O, et al. Unsupervised word embeddings capture expert-designed materials descriptors. Nature. 2019; 571: 95-100.
  16. Fleming N. How artificial intelligence is changing drug discovery. Nature. 2018; 557: 55-57.
  17. Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021; 596: 583-589.
  18. Burger B. A mobile robotic chemist. Nature. 2020; 583: 237-241.
  19. Esteva, A, Kuprel B, Novoa RA, Swerrer SM, Blau HM, Thrun S, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017; 542: 115-118.
  20. Topol E J. High-performance medicine: the convergence of human and artificial intelligence. Nature Medicine. 2019; 25: 44-56.
  21. Vaidya S. Artificial intelligence and machine learning in public health surveillance. JMIR Public Health and Surveillance. 2020; 6: 19771.
  22. Mak KK, Pichika MR. Artificial intelligence in drug development: current status and future prospects. Drug Discovery Today. 2019; 24: 773-780.
  23. Mettler T. From surgical robots to robotic care: A review of artificial intelligence applications in healthcare. Technological Forecasting and Social Change. 2014; 80: 1146-1156.
  24. Skolovy M. Generative design for additive manufacturing: a review. Additive Manufacturing. 2020; 35: 101429.
  25. Lee J, Bagheri B, Kao HA. A cyber-physical systems architecture for Industry 4.0-based manufacturing systems. Manufacturing Letters. 2013; 1: 1-6.
  26. Bonin-Font F, Ortiz-Monasterio A, C Valavanis K. Recent Trends and Advances in Robotics and Automation. Robotics and Autonomous Systems, 2019; 120: 103-233.
  27. Mohanty S. A Comprehensive Survey on Using AI in Smart Cities. IEEE Access. 2018; 6: 25297-25313.
  28. Vlachos AG, Koulocheris DN. Artificial intelligence in aerospace: A review of applications. Journal of Aerospace Engineering. 2018; 31: 04018080.
  29. Waldchen J, Mader P. Machine learning for image-based plant phenotyping. Plant Methods. 2018; 14: 70.
  30. Kumar A, Singh PK. A survey on air pollution prediction using machine learning techniques. Environmental Modeling and Software. 2019; 118: 195-207.
  31. Nesmachnow S. Deep learning for smart grid applications. Smart grid and renewable energy. 2017; 8: 397.
  32. Caron E. Artificial intelligence in high energy physics. Journal of Instrumentation. 2012; 7: C10014.
  33. Zuo W. Deep learning for gravitational wave detection and parameter estimation. Natural Astronomy. 2020; 4: 1084-1090.
  34. Libbrecht MW, Noble WS. Machine learning applications in genetics and genomics. Nature Reviews Genetics. 2015; 16: 321-332.
  35. King RD, Rowland J, Oliver SG, Young M, Aubrey W, Byrne E, et al. The automation of science. Science. 2009; 324: 85-89.
  36. Trewyn RW, Chen Y. AI-driven hypothesis generation for drug repurposing. AI Magazine. 2019; 40: 48-58.
  37. Zhavoronkov A, Ivanenkov YA, Aliper A, Veselov MS, Aladinskiy VA, Aladinskaya AV, et al. Deep learning enables rapid identification of potent DDR1 kinase inhibitors. Nature Biotechnology. 2019; 37: 1138-1144.
  38. Cohan A, Feldman S, Beltagy I, Downey D, Weld DS. SPECTER: Document-level Representation Learning using Citation-informed Transformers. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020; 2137-2144.
  39. Cui A. Knowledge graph based drug repurposing: a review. Briefings in Bioinformatics. 2020; 21: 2095-2115.
  40. Jensen KF. The flow chemistry revolution. Journal of Flow Chemistry. 2017; 7: 1-10.
  41. Grieves M, Vickers J. Digital Twin: Mitigating unpredictable, undesirable emergent behavior in complex systems. Transdisciplinary Perspectives on Complex Systems. 2017; 85-113.
  42. Obermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting racial bias in an algorithm used to manage the health of populations. Science. 2019; 366: 447-453.
  43. Adadi A, Berrada M . Peeking inside the black-box: A survey on Explainable Artificial Intelligence (XAI). IEEE Access. 2018; 6: 52138-52160.
  44. Frey CB, Osborne MA. The future of employment: How susceptible are jobs to computerisation?. Technological Forecasting and Social Change. 2017; 114: 254-280.
  45. Ziewitz M. Governing Algorithms: Myth, Message, and Methods. Science, Technology, and Human Values. 2016; 41: 3-16.
  46. Strubell E, Ganesh A, McCallum A. Energy and policy considerations for deep learning in NLP. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. 2019; 3645-3651.
  47. Lake BM, Ullman TD, Tenenbaum JB, Gershman SJ. Building machines that learn and think like people. Behavioral and Brain Sciences. 2017; 40.
  48. Dunjko V, Briegel HJ. Machine learning and artificial intelligence in the quantum domain. Reports on Progress in Physics. 2019; 81: 074001.
  49. Garcez AS, Lamb LC. Neurosymbolic AI: The 3rd Wave. Artificial Intelligence Review. 2020; 54: 3121-3132.
  50. McMahan HB, Moore E,  Ramage D, Hampson H,  Arcaet BAY . Communication-Efficient Learning of Deep Networks from Decentralized Data. Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. 2017; 1273-1282.
  51. AlQuraishi M. AlphaFold at CASP13. Bioinformatics. 2019; 35: 4862-4865.
  52. Bjornsson L.The digital twin in clinical medicine. NPJ Digital Medicine. 2020; 3:1-8.
  53. Caliskan A. Semantics derived automatically from language corpora contain human-like biases. Science. 2017; 356: 183-186.
  54. Gulshan V, PengL, Coram M, Stumpe MC, Wu  D, Narayanaswamy  A, et al. Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs. JAMA. 2016; 316: 2402-2410.
  55. Schwaller P. Prediction of retrosynthetic reactions using neural sequence-to-sequence models. ACS Central Science. 2019; 5: 1543-1552.
  56. Ulissi ZW, Zhong M, Tran K, Mi Y, Wang C, Wang Z, et al. Accelerated discovery of new electrocatalysts using active machine learning. Nature Catalysis. 2017; 1: 74-81.