Integration of AI and Robotics in Smart Manufacturing: A Case Study Approach

Integration of AI and Robotics in Smart Manufacturing: A Case Study Approach

1Gurwinder Kaur , 2Sachin Yadav

Department of Computer Application 1Chandigarh School of Business Jhanjeri , Mohali 2Chandigarh Group of Colleges Jhanjeri, Mohali 1gurwin2002@gmail.com , 2sachinyadav170@gmail.com

Abstract — Robotics, the amalgamation of perception and action, inherently relies on Artificial Intelligence (AI) to imbue these interactions with intelligence. This paper explores the pivotal role of AI in robotics, emphasizing the challenges posed by real-world applications and the need for sophisticated cognitive abilities. While AI addresses fundamental questions of knowledge representation and analysis, robotics compels AI to grapple with the complexities of physical environments and object manipulation. This paper reviews AI contributions to perception and object-oriented reasoning, illustrating the intersection of these disciplines through practical examples.

Keywords— Robotics, Artificial Intelligence, perception, object- oriented reasoning, cognition, real-world applications.

1. INTRODUCTION

   The paper's introduction provides an in-depth history of the growth of robotics It highlights the vital role that machine learning (AI) plays as the software agent in charge of robotic gadgets, emphasizing the beneficial link between robotics and AI. Transitioning to the theoretical and practical aspects, the paper delves into robot hardware, with a specific focus on sensors serving as the perceptual interface between robots and their environment. It differentiates between passive sensors like cameras and active

   sensors such as sonar, radar, and laser, elucidating their functions and applications. Additionally, the discussion extends to imaging sensors and contraceptive sensors, which provide insights into the robot's internal state.

   Figure 1: The role of artificial intelligence in Robotics

   1. About

      The section on effector elaborates on how robots interact with the environment and alter their body shape, introducing the concept of degrees of freedom (DOFs) to quantify a robot's independent movement capability. It differentiates between manipulators and mobile robots based on their degrees of freedom and accountability.

      In the subsection focusing on movement, various locomotion mechanisms for mobile robots, such as differential drive and synchronic drive, are explored, along with their advantages and limitations. Additionally, power sources for robots, primarily electric motors, are discussed, with brief mentions of alternative mechanisms like pneumatic and hydraulic actuation.

      The paper also touches upon digital communication modules and underscores the significance of body frames in robot construction. It concludes with a discussion on robotic perception, highlighting the challenges posed by noisy sensor data and dynamic environments, and the importance of developing effective internal representations for robots.

      The narrative then delves into early conceptions of robots, citing historical instances of mechanical devices designed for specific physical tasks dating back to ancient times. It discusses various automates created in different periods, leading up to the 20th-century advancements in robotics.

      Finally, it highlights the emergence of modern robots in the 1950s, notably with George C. Devil's invention of the re- programmable manipulator "Unimate" and Joseph Challenger's role in commercializing industrial robots. Academic endeavours, such as the development of "Shakes" at the Stanford Research Institute, are also acknowledged for their contributions to robotic advancements.

      Figure 2: History of Robotics

   2. Challenges in Robotics and AI

      1. Complexity of Real-world Environments:Robotics and AI face the significant challenge of developing systems capable of effectively navigating and operating in diverse and dynamic real-world environments. These environments often introduce uncertainties like varying lighting conditions, obstacles, and unpredictable events, necessitating robots to adapt and make real-time decisions.

      2. Sensor Limitations and Perception: Sensors are pivotal in enabling robots to perceive their surroundings, yet they have limitations in accuracy, range, and reliability. Overcoming these limitations and creating robust perception systems capable of accurately interpreting sensor data in complex environments remains a substantial challenge.

      3. Autonomous Decision-Making: Achieving high-level autonomy in robotic systems entails developing algorithms and decision-making frameworks that empower robots to make complex decisions independently. Tasks such as path planning, object recognition, and action selection require sophisticated AI techniques like machine learning and planning.

      4. Safety and Reliability:Ensuring the safety and reliability of robotic systems is critical, especially in scenarios where robots interact with humans or operate in hazardous environments. Robust safety mechanisms, fail-safe systems, and ethical considerations are essential to prevent accidents and ensure responsible deployment of AI-powered robots.

      5. Human-Robot Interaction: Designing robots capable of effectively interacting and collaborating with humans presents a multifaceted challenge. This involves developing intuitive interfaces, natural language processing capabilities, and social skills to facilitate seamless communication and cooperation between humans and robots across various contexts.

      6. Data Privacy and Security:AI-driven robotic systems rely on vast amounts of data for training and operation, raising concerns about data privacy and security. Safeguarding sensitive data and protecting against cyber threats and malicious attacks is crucial for maintaining the integrity and trustworthiness of robotic systems.

      7. Interdisciplinary Collaboration:Robotics and AI span various disciplines, including computer science, engineering, psychology, and ethics. Promoting interdisciplinary collaboration and knowledge exchange is essential for advancing research and tackling the complex challenges inherent in the field.1.3

   3. Applications of Robotics and Artificial Intelligence

      1. Manufacturing and Industrial Automation:

         Robotics and AI find extensive applications in manufacturing for tasks like assembly, welding, painting, and material handling.

         AI-driven robotic systems optimize production processes, enhance quality control, and boost productivity in factory settings.

      2. Healthcare:

         Robots assist in surgical procedures, including minimally invasive surgery, precision surgery, and rehabilitation.

         AI-powered medical imaging techniques aid in diagnosis, treatment planning, and disease monitoring.

         Robotics and AI support patient care, including mobility assistance through robotic exoskeletons and elderly care through robotic companions.

      3. Logistics and Supply Chain Management:

         Autonomous mobile robots streamline warehouse operations, including sorting, inventory management, and order fulfilment.

         AI algorithms optimize logistics operations, route planning, and scheduling for transportation and distribution networks.

      4. Agriculture:

         Agricultural robots automate tasks such as planting, harvesting, weeding, and spraying, leading to increased efficiency and reduced labour costs.

         AI algorithms analyze data from sensors and drones to optimize crop management practices, improve yield prediction, and monitor plant health.

      5. Autonomous Vehicles and Transportation:

         Self-driving cars and autonomous drones utilize AI for perception, navigation, and decision-making, promising safer and more efficient transportation.

         AI-powered traffic management systems optimize traffic flow, reduce congestion, and enhance road safety in urban environments.

      6. Retail and Customer Service:

         Robots automate inventory management, shelf stocking, and customer assistance in retail stores, thereby improving operational efficiency and customer experience.

         AI-driven chatbots and virtual assistants offer personalized recommendations, address customer inquiries, and facilitate online transactions.

      7. Finance and Banking:

         AI algorithms analyze financial data to detect fraud, assess credit risk, and improve decision-making processes in banking and finance.

         Robotic process automation (RPA) automates repetitive tasks like data entry, document processing, and account reconciliation, thus enhancing operational efficiency.

      8. Education and Training:

         Educational robots and AI-powered tutoring systems facilitate personalized learning experiences by offering interactive lessons and adaptive feedback to students.

         Virtual reality (VR) and augmented reality (AR) technologies create immersive training environments for skill development and professional training across various industries.

2. Literature Review

   1. Author(s): Lee, Longueuil et Al.

      Year: 2017

      Title: "Challenges and Opportunities of Artificial Intelligence in Robotics"

      Summary: This paper discusses the challenges and opportunities of integrating AI in robotics, addressing issues such as handling uncertain and dynamic environments, safety, robustness, and interpretability, while also exploring potential applications and future directions. Artificial intelligence has become a major technological development that has the potential to revolutionize the robotics industry. Recent years have witnessed a burgeoning desire to integrate AI techniques into robotic systems, empoweringtheir ability to work alone and adjust to shifting conditions.

      Robotics has made substantial use of machine learning, which includes supervised, unsupervised, and reinforcement learning approaches, to enable robots to learn from data and make decisions based on patterns and experiences. Concurrently machine vision, a subset of artificial intelligence, has significantly bolstered robotic perceptual abilities, which allow robots to use visual clues to understand and navigate their environment. Additionally, natural language processinghas been harnessed To make things easier and seamless human- robot communication,allowing robots to comprehend and produce spoken language for effective dialogue and cooperation across various applications, including personal assistance, service, and social robots.

   2. Author(s): Murphy, Robin R.

      Year: 2000

      Title: "Introduction to AI Robotics"

      Summary: This introductory text provides an overview of AI techniques in robotics, covering topics such as machine learning, computer vision, and natural language processing, offering insights into the integration of AI in robotic systems.

      Deep learning, a subfield of machine learning, characterized by training artificial neural networks with vast datasets, has showcased exceptional performance across numerous robotics tasks, spanning from from motion planning and speech recognition to object recognition. reinforcement of learning, another variant of artificial intelligence, has been instrumental in training Robots to learn how to make the best decisionstrategies via iterative relationships with their surroundings. Notwithstanding the scientificstrides made, The ethical implications of artificial intelligence in robots have acquired prominence, necessitating a focus on safety, transparency, accountability, and bias mitigation to ensure ethical deployment across diverse domains.

      Despite the rapid advancements, challenges persist, such as the restrictions placed on artificial intelligence algorithms in navigating surroundings that are dynamic and uncertain, alongside concerns connected to protection, hardiness, and Readability. Furthermore, The broader societal and financial consequences of broad Robotics' usage of AI, including possible impacts regarding jobs and social standards, require

      meticulous consideration. However, the possible uses of artificial intelligence in robots.

      However, the possible uses of artificial intelligence in robots are expansive, Using the fusion in robotics and AI poised to proceed with reshaping automation, thereby presenting both Prospects and obstacles for scholars, engineers, decision- makers, and the community at large.

3. Research Methodology

   The methodology employed in this study was carefully selected to ensure a comprehensive understanding of the current state of knowledge in the field of Artificial Intelligence (AI) in Management Information Systems (MIS). Drawing from established research traditions, the literature review served as a pivotal tool for comprehending the latest advancements and identifying critical knowledge gaps within the domain (Scorecard et Al., 2006). This approach not only aids in understanding what is currently known but also sheds light on areas requiring further exploration, thereby motivating researchers to bridge existing gaps (Webster & Watson, 2002).

   The chosen methodology for this study was a state-of-the-art review, aimed at investigating the current landscape of knowledge, pinpointing areas for future research, identifying research limitations, and offering fresh perspectives while remaining firmly rooted in existing literature (Booth, Sutton, & Announced, 2016). Following an organizational review framework proposed by Webster and Watson (2002), a concept-eccentric approach was adopted. A formal systematic literature review protocol, guided by recommendations from Levy and Ellis (2006), Vom Brocke et al. (2009), Webster and Watson (2002), and Mathiassen et al. (2012), was adhered to for extracting, analyzing, interpreting, and reporting literature- based findings.

   The research commenced with defining a search strategy, which involved selecting appropriate search terms derived from the research questions of the study. These terms were

   systematically employed in database searches of selected journals, guided by the Association for Information Systems (AIS) journal basket list (Lowery et al., 2013). After retrieving relevant papers, abstracts, titles, and keywords were scrutinized to ensure alignment with the study's objectives. Papers not related to the research scope were excluded, along with those where AI played a peripheral role. Following a rigorous screening process, a subset of 74 papers was deemed pertinent for detailed analysis.

   To focus the study and achieve the research goals effectively, specific research questions were formulated based on prior works (Scornavacca et al., 2004, 2006). These questions aimed to elucidate various aspects such as the contextual domain of the study, research methods employed, data collection strategies, the role of AI, subfields studied, techniques used, and research agendas proposed by authors. Each

   paper was systematically coded to answer these questions, facilitating a clear understanding of the research landscape in AI within the MIS field.

   Overall, the methodology adopted in this study ensured a rigorous and systematic approach to reviewing the literature, providing valuable insights into the current state of knowledge and paving the way for future research end favours.

4. Conclusion and future scope

   Artificial Intelligence (AI) and robotics offer immense potential for societal transformation, yet they also pose ethical dilemmas, including concerns regarding abuse, independence, prejudices, openness, and their influence on human relationships and employment. To tackle certain difficulties, it is imperative to integrate ethical considerations throughout the design phase. This involves recognizing and eliminating biases, maintaining openness and human oversight, additionally fostering robotic ethics that align with the principles of humanity.

   Finally, several AI algorithms used for robotics have distinct advantages and disadvantages. Although supervised learning is faster than reinforcement learning in flexible decision-making is advantageous Regarding tasks with Data labelling, machine vision enhances With its capacity for visual perception, SLAM is vital Evolutionary algorithms work well for navigation and mapping, and deep learning is useful for perception and control, and optimization. The choice algorithm is dependent upon factors such as task specifications, data accessibility, computing power, and the robot's intended level of autonomy. The future scope of robotics and AI is vast and holds tremendous potential for transforming various aspects of society. Some key areas where robotics and AI are expected to make significant advancements in the future include:

   1. Autonomous Vehicles: The development of self-driving cars and drones is expected to revolutionize transportation and logistics, leading to safer and more efficient mobility solutions.

   2. Healthcare: Robotics and AI are poised to revolutionize healthcare by enabling the development of robotic surgeons, personalized medicine, remote patient monitoring, and assistive devices for the elderly and disabled.

   3. Manufacturing: The use of robotics and AI in manufacturing processes is expected to lead to increased automation, improved efficiency, and customization capabilities, thereby transforming traditional manufacturing industries.

   4.4 Smart Cities:Robotics and AI innovations have a significant potential impact on building smart cities by enabling efficient resource management, intelligent infrastructure, and responsive urban planning.

   4.5.Agriculture: Robotics and AI have the potential to revolutionize agriculture through precision farming techniques, autonomous farming equipment, and real-time monitoring of crops and livestock.

5. References

1. F. Ingrand, M. Ghallab

Deliberation for autonomous robots: a survey Artif. Intell., 247 (2017), pp. 10-44

1. T. Hester, P. Stone

   Intrinsically motivated model learning for developing curious robots Artif. Intell., 247 (2017), pp. 170-186

2. K. Ramirez-Amaro, M. Beetz, G. Cheng

   Transferring skills to humanoid robots by extracting semantic representations from observations of human activities

   Artif. Intell., 247 (2017), pp. 95-118

3. L. Kunze, M. Beetz

   Envisioning the qualitative effects of robot manipulation actions using simulation-based projections

   Artif. Intell., 247 (2017), pp. 352-380

4. J. Pajarinen, V. Kyrki

   Robotic manipulation of multiple objects as a POMDP Artif. Intell., 247 (2017), pp. 213-228

5. M. Tenorth, M. Beetz

   Representations for robot knowledge in the KnowRob framework Artif. Intell., 247 (2017), pp. 151-169

6. V.R. Kompella, M. Stollenga, M. Luciw, J. Schmidhuber

   Continual curiosity-driven skill acquisition from high-dimensional video inputs for humanoid robots

   Artif. Intell., 247 (2017), pp. 313-335

7. Z. Wang, A. Boularias, K. Mülling, B. Schölkopf, J. Peters Anticipatory action selection for human-robot table tennis Artif. Intell., 247 (2017), pp. 399-414

8. A. Agostini, C. Torras, F. Wörgötter

   Efficient interactive decision-making framework for robotic applications

   Artif. Intell., 247 (2017), pp. 187-212

9. D. Martínez, G. Alenyà, C. Torras

   Relational reinforcement learning with guided demonstrations Artif. Intell., 247 (2017), pp. 295-312

10. S. Lemaignan, M. Warnier, E.A. Sisbot, A. Clodic, R. Alami

    Artificial cognition for social human-robot interaction: an implementation

    Artif. Intell., 247 (2017), pp. 45-69

11. J. Bidot, L. Karlsson, F. Lagriffoul, A. Saffiotti

    Geometric backtracking for combined task and motion planning in robotic systems

    Artif. Intell., 247 (2017), pp. 229-265

12. T.T. Nguyen, T. Silander, Z. Li, T.-Y. Leong

    Scalable transfer learning in heterogeneous, dynamic environments Artif. Intell., 247 (2017), pp. 70-94

13. M. Günther, T. Wiemann, S. Albrecht, J. Hertzberg

    Model-based furniture recognition for building semantic object maps

    Artif. Intell., 247 (2017), pp. 336-351 Marc Hanheide, et al.

    Robot task planning and explanation in open and uncertain worlds Artif. Intell., 247 (2017), pp. 119-150

14. A. Faust, I. Palunko, P. Cruz, R. Fierro, L. Tapia

    Automated aerial suspended cargo delivery through reinforcement learning

    Artif. Intell., 247 (2017), pp. 381-398

15. A.G. Hofmann, B. Williams

    Temporally and spatially flexible plan execution for dynamic hybrid systems

    Artif. Intell., 247 (2017), pp. 266-294

16. A. Kupcsik, M.P. Deisenroth, J. Peters, L.A. Poh, P. Vadakkepat, G. Neumann

    Model-based contextual policy search for data-efficient generalization of robot skills

    Artif. Intell., 247 (2017), pp. 415-439

17. E.A. Feigenbaum, J. Feldman, et al.

    Computers and Thought AAAI Press (1963)

    International Federation of Robotics World Robotics 2016

    SPARC The Partnership for Robotics in Europe

    Strategic Research Agenda for Robotics in Europe 20142020 Robotics VO

18. A Roadmap for U.S. Robotics

    N. Wiener

    Cybernetics: Control and Communication in the Animal and the Machine

    Wiley, New York (1948)

19. J. Needham

    Science and Civilisation in China, vol. 2

    History of Scientific Thought, Cambridge University Press (1954)

20. N.J. Nilsson

    Shakey the Robot Tech. Rep.

    SRI Artificial Intelligence Center (1984)

21. M. Ghallab, D. Nau, P. Traverso Automated Planning Theory and Practice Elsevier Science (2004)

22. S. LaValle

    Planning Algorithms

    Cambridge University Press (2006)

23. T. Dean, M. Wellman Planning and Control

    Morgan Kaufmann Publishers Inc. (1991)

24. K. Wagstaff

    Machine learning that matters

    Proceedings of 29th International Conference on Machine Learning, Edinburgh, Scotland (2012)

25. A. Saffiotti, K. Konolige, E.H. Ruspini

A multivalued logic approach to integrating planning and control Artif. Intell., 76 (12) (1995), pp. 481-526