Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 945 https://internationalpubls.com The P versus NP Problem Insights into Computational Complexity 1S. Balamuralitharan, 2R. Arulprakasam, 3N.Sujatha, 4M Bala Prabhakar, 1Adjunct Faculty, Department of Pure and Applied Mathematics, Saveetha School of Engineering, SIMATS, Chennai, Tamil Nadu, India Email Id: balamurali.maths@gmail.com 2Department of Mathematics, College of Engineering and Technology, SRM Institute of Science and Technology, SRM Nagar, Kattankulathur - 603203, Chengalpattu District, Tamilnadu, India r.aruljeeva@gmail.com 3Associate Professor, Department of Mathematics, Aditya University, Surampalem, India, sujatha.nanduri@aec.edu.in 4Associate Professor, Dept of Mathematics, Aditya University, Surampalem, India, balaprabhakar.mattaparthi@aec.edu.in Article History: Received: 12-10-2024 Revised: 15-11-2024 Accepted: 19-12-2024 Abstract: Theoretical computer science along with mathematics face an ultimate problem which remains unsettled because of its deep significance as P versus NP. The problem seeks to know if all the problems that can be validated quickly in polynomial time can also be solved quickly. Researchers have spent multiple decades exploring this question alongside significant practical applications yet the issue has not been resolved. This paper delivers an efficient exploration of the P vs. NP problem by studying its core components in computational complexity theory while analyzing historical and contemporary studies alongside methodological frameworks for solving this problem. The research identifies outcomes together with their interpretive values and implications to offer extensive comprehension between technical aspects and philosophical evaluations of the problem which enables advanced understanding of algorithm development and cryptographic practices along with optimization techniques and artificial intelligence methodologies. Keywords— P vs. NP, Computational Complexity, Polynomial Time, NP-completeness, Algorithms, Cryptography, Theoretical Computer Science I. INTRODUCTION The main question of the problem concerns whether every checking problem that uses polynomial time verification also has polynomial time solutions. The P versus NP question explores whether every polynomial time solvable problem by a deterministic Turing machine also features a polynomial time checking capability when using a nondeterministic Turing machine. Both computer science and the world at large explore this central subject after Stephen Cook formally introduced the question in 1971 through “The Complexity of Theorem- Proving Procedures [1-2]. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 946 https://internationalpubls.com The answer to this question would affect fundamental computer science inquiries and theoretical research. The solution of P = NP would enable computers to address several hard problems including Traveling Salesman Problem (TSP), Integer Factorization along with Boolean Satisfiability Problem (SAT) with high operational effectiveness. The solution of P = NP would transform optimization along with cryptography machine learning and artificial intelligence fields. The discovery of P ≠ NP would create essential limits to what computers can efficiently compute which would provide better boundaries between complicated and simple problems. It becomes difficult to establish either type of proof because of the challenge. Thousands of mathematicians have looked for polynomial-time algorithms to solve any of the NP-complete problems without success since these problems stand among the toughest to solve within the NP class [11]. According to the Cook-Levin Theorem (1971) SAT was proven to be NP- complete thus demonstrating that solving any NP-complete problem in polynomial time would lead to the ability to solve all problems within the NP class in polynomial time also. Besides SAT the concept now extends to various additional problems including Knapsack along with Clique and Vertex Cover. RSA encryption together with other cryptographic systems relies on the fundamental concept of P versus NP to protect critical data. This problem stands as the foundation behind cryptographic encryption because it remains unproven whether particular problems require long periods to solve. The discovery of P = NP would allow quick solutions to break encryption methods thereby requiring a rapid development of new cryptographic methods. The belief that some problems (including the training of large neural networks) cannot be solved within polynomial time functions as fundamental base for AI along with machine learning and optimization algorithms which determine heuristic design [14-15]. Research activities about this unsolved problem have been substantial during the last fifty years although a solution has not been found. Due to P vs. NP most researchers study the structure of NP-complete problems together with searching for polynomial-time solutions for these problems. Toolmakers developed approximation algorithms coupled with randomized algorithms and heuristic algorithms to produce workable solutions for many practical situations although such systems ensure only approximate results. The scientific community strongly believes P is not equal to NP yet no final verification exists which continues to fascinate math and computer science researchers. Many consider this core question important enough for the Clay Mathematics Institute to select it as one of seven Millennium Prize Problems with a $1 million prize reward for its successful resolution [12]. The relationship between P and NP would indicate the existence of a hidden practical computational approach if P = NP turns out to be true. If P ≠ NP then researchers would gain further confirmation that human computation has inherent restrictions. The solution of this Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 947 https://internationalpubls.com mathematical problem will bring scientific progress to theoretical computer science and reshape our comprehension of what the computational universe entails. Novelty and Contribution The research intends to conduct a detailed analysis of the P versus NP problem by tracing its historical progression together with its practical effects and active resolution activities. The novel aspect of this research combines theoretical perspectives from computer science with practical insights to study the problem [8]. This paper conducts a thorough assessment of contemporary advancements in complexity theory around NP-complete issues. The paper connects theoretical analysis of P vs. NP with practical applications which include cryptography together with machine learning optimization and artificial intelligence. This paper uses theoretical problem background together with practical effects to offer improved comprehension of P vs. NP solutions for various fields. The paper presents a fresh framework that deals with the problem. The article investigates computational models beyond Turing machines by studying quantum computing and probabilistic computation to obtain fresh perspectives about P vs. NP. New computational models emerging today create distinct pathways to solve the P = NP question through P = NP evidence or evidence that P is not equal to NP. The paper explores fundamental philosophical questions associated with the problem which frequently escape attention during technical P vs. NP talks. The research delves into philosophical areas of the P vs. NP problem to support a wider examination of computation nature and human knowledge boundaries together with advances in fundamental properties of computational processes [10]. The paper provides a discussion on prospective research pathways for the P vs. NP problem together with open questions and existing conjectures which could potentially lead to breakthrough solutions. The research investigates the present situation of this problem point- by-point along with establishing future development paths towards progress. II. RELATED WORKS In 2022 D. Bianco et al., [13] introduced the P versus NP problem became the main focus of computational complexity theory research when it emerged and caused extensive academic discussions about efficient computing. Scientists pursue this topic because they seek to identify the limits of efficient computation. The main obstacles in problem classification involve separating P from NP as P contains problems that require polynomial time solutions while NP includes problems where polynomial time verification exists. During the early period researchers examined NP-complete problems to understand their traits since these problems represent the most challenging problems in the NP category. The Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 948 https://internationalpubls.com significance of NP-complete problems lies in their ability to solve every problem in NP if a single NP-complete problem receives polynomial-time solution thus equating P to NP. Scientists during early research identified fresh NP-complete challenges and examined these problems extensively. NP-complete problems proved to be a wide-ranging group that exists across multiple subject domains including logical analysis besides scheduling and network planning and optimization techniques [9]. In 2022 D. Gamarnik et.al., C. Moore et.al., and L. Zdeborová et.al., [3] suggested the scientists have extensively studied approximation algorithms and heuristic methods as solutions for NP-complete problems when used in practical applications. Research focuses on developing approximation algorithms that generate acceptable solutions within specific time limits because the exact solution of these problems demonstrates unlikely polynomial-time feasibility under P ≠ NP assumptions. Such solution approaches hold great importance in optimization applications because they yield valuable close-to-optimal results for tasks including resource management and route design and artificial intelligence models. The relationship between P and NP has been studied using different complexity classes which exist between P and NP. The research community continues to analyze the boundaries which define NP-hard and NP-easy categories even though these boundaries lie between P and NP. Problem classification under these subcategories determines the difficulty level of unverified and unsolvable problems that do not belong to NP-complete. Research about computational theory depends heavily on our comprehension of these labeled problem classes. The P vs. NP question leads theoretical cryptography to experience important developments. The security of many encryption systems depends on hard problems that need polynomial time resolution to maintain their strength after validating whether P equals NP. The equivalence of P and NP would degrade the security of cryptographic systems thus leading to the need for new cryptographic protocols altogether. In 2020 N. Almasarwah et.al. and G. A. Süer et.al., [7] proposed the research focusing on philosophical aspects regarding P vs NP problem theory has become increasingly prevalent during the last few years. Numerous philosophers along with theoretical experts continue to discuss whether current computational methods are sufficient for resolving this problem or if an alternative computing approach must be developed. Discussions explore knowledge definitions and human creative potential since they involve understanding how people think when handling complex brain or artificial intelligence systems. Quantum computing developments represent the newest addition to P vs. NP research. Quantum computing technology demonstrates ability to tackle specific problems with exponential speed relative to traditional computing systems thus creating new perspectives regarding P vs. NP problem resolution. Research on quantum computing continues to establish its potential effect on computational complexity which could create either enhanced or barriers in the understanding of P and NP. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 949 https://internationalpubls.com These areas of research collectively expand understanding about the P versus NP problem by demonstrating both the complications and solutions which could transform both computer science and numerous practical fields that require speedy digital processing. III. PROPOSED METHODOLOGY Computational complexity theory deals with the fundamental challenge presented by P versus NP problem. The proposed paper introduces a new method to investigate the problem through different computational algorithms and theoretical analysis of models along with algorithmic strategies. The methodology uses theoretical examinations which are combined with experimental tests to research the relationship possibilities between P and NP. The goal consists of establishing whether P and NP are equivalent powers and exploring practical effects on current computational systems [4]. The reduction method demonstrates the creation of links between two problems so that tackling the first could generate solutions to the second within polynomial time. Establishing these reductions enables us to understand NP-complete problems better and determine their relationship to P. The focus of our investigation involves determining if Boolean satisfiability (SAT) can be solved within polynomial time. When SAT demonstrates reducibility to a problem which has polynomial-time solvability it necessarily confirms P=NP. First we analyze how NP-complete problems are constructed but then we explore whether a polynomial-time algorithm exists for solving these problems. The proposed algorithm for any NP-complete problem can serve other NP problems because reduction properties maintain transitivity. The systematic method establishes procedures for building verification algorithms which check solutions efficiently while investigating polynomial-time solutions for NP-complete problems [5]. The proposed method requires these mathematical expressions to work. The equations present multiple computational models that serve as both decision functions and to confirm complex problem verification through reductions. In this workplace we first establish decision problems as tasks that need to verify if solutions exist inside specific computational models. The decision problem X contains its solution space defined as S, Where: 𝑆 = {𝑥 ∣ 𝑓(𝑥) = 1} Here, 𝑓(𝑥) is the decision function for the problem 𝑋, returning a binary outcome. For a problem in 𝑁𝑃, the function 𝑓(𝑥) must be verifiable in polynomial time. In order to model reduction between problems, we use the concept of polynomial-time transformations. Let 𝐴 be a problem in NP and 𝐵 a problem in P. A reduction from 𝐴 to 𝐵 can be defined as a function 𝑅 such that: Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 950 https://internationalpubls.com 𝐴(𝑥) ≤𝑃 𝐵(𝑅(𝑥)) where ≤𝑃 represents polynomial-time reducibility. If 𝐴 can be reduced to 𝐵 in polynomial time, solving 𝐴 becomes equivalent to solving 𝐵. Next, we define polynomial-time algorithms that verify solutions for NP problems. For any problem 𝑋 ∈ 𝑁𝑃, there exists a verification function 𝑉(𝑥, 𝑦) where 𝑦 is a proposed solution and 𝑉(𝑥, 𝑦) checks whether 𝒚 satisfies the conditions of 𝑋. The time complexity of the verification function is denoted as: 𝑇𝑉(𝑛) = 𝑂(𝑛𝑘) where 𝑘 is the degree of the polynomial. This polynomial-time verification is a key characteristic of NP problems. Consider the Boolean satisfiability problem (SAT). The Boolean formula 𝜙 in SAT can be represented as: 𝜙 = (𝑥1 ∨ 𝑥2 ∨ ⋯∨ 𝑥𝑛) where V represents the logical OR operation, and 𝑥𝑖 are Boolean variables. A solution 𝛼 is a truth assignment to the variables 𝑥1, 𝑥2, … , 𝑥𝑛, and we want to check whether there exists an assignment such that 𝜙 evaluates to true. The problem is NP-complete, and if SAT can be solved in polynomial time, then all problems in NP can also be solved in polynomial time. To analyze the complexity of NP problems, we define the complexity class of a problem 𝑋 as: Complexity(𝑋) = min(𝑇(𝑛) ∣ ∃𝑦 such that 𝑉(𝑥, 𝑦) runs in 𝑂(𝑛𝑘)) where 𝑇(𝑛) represents the time required to solve the problem. If Complexity(𝑋) = 𝑂(𝑛𝑘) for some polynomial 𝑘, then 𝑋 belongs to NP. We also explore the role of probabilistic algorithms in solving NP problems. A probabilistic Turing machine (PTM) has access to random bits and operates with a probability of error. Let 𝑃 be a probabilistic algorithm and 𝑋 an NP problem, where the time complexity of 𝑃 is denoted by: 𝑇𝑃(𝑛) = 𝑂(𝑛𝑘 ⋅ log⁡ 𝑛) This approach explores the possibility of using randomness to solve NP problems more efficiently. The relationship between 𝐏 and 𝐍𝐏 is further explored through the circuit complexity of Boolean functions. The circuit size of a Boolean function 𝑓 is defined as the smallest number of gates required to compute 𝑓, and is given by: Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 951 https://internationalpubls.com CircuitSize(𝑓) = min (∑   𝑛 𝑖=1  𝑔𝑖 ∣ 𝑔𝑖 ∈ gates used in the circuit ) By analyzing the circuit complexity of NP-complete problems, we aim to identify structural patterns that may lead to polynomial-time algorithms or further evidence of the separation between P and NP. Finally, the search space for finding polynomial-time solutions to NP-complete problems is examined using dynamic programming. Given a problem with a solution space of size 2𝑛, a dynamic programming approach breaks down the problem into subproblems with overlapping substructure, reducing the overall complexity. The dynamic programming recurrence can be written as: 𝑑𝑝(𝑖, 𝑗) = min(𝑑𝑝(𝑖 − 1, 𝑗), 𝑑𝑝(𝑖 − 1, 𝑗 − 1) + cost(𝑖)) where 𝑑𝑝(𝑖, 𝑗) represents the minimal cost of solving the problem at step 𝑖 with state 𝑗, and cost(𝑖) is the cost at step 𝑖. FIGURE 1: FLOWCHART OF METHODOLOGY FOR EVALUATING P VS NP THROUGH ALGORITHMIC ANALYSIS Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 952 https://internationalpubls.com This methodology sets the stage for deeper exploration into the P versus NP problem, providing a structured approach for both theoretical exploration and empirical testing. By utilizing both existing models and introducing new computational frameworks, the research aims to shed light on the central question in computational complexity theory. IV. RESULT & DISCUSSIONS Our methodology produces result about computational complexity theory which investigate potential P and NP relationships. The research employed data simulations together with theoretical concepts for evaluating its performance. An analysis of experimental findings follows in this section while performance results get evaluated and the current P versus NP debate gets investigated [6]. We examined NP-complete computational challenges through dynamic programming using the TSP and Knapsack Problem as main research targets. The recognized problems from computational theory create perfect conditions to evaluate new algorithms. Dynamic programming created substantial improvements to time complexity but the solution duration for large datasets demonstrated exponential growth patterns. The algorithm operated with polynomial efficiency during modest problem executions although its performance severely declined when problems grew larger. The findings were confirmed by establishing execution times versus variable numbers and analyzing them relative to theoretical predictions reported earlier [7]. The main outcome from our testing revealed how NP-complete problems resist polynomial-time solvers beyond specific difficulty thresholds. Dynamic programming in the TSP first achieved faster computation but expanding city numbers caused the path optimization time to explode exponentially. The table shows how dynamic programming performances better than traditional brute-force solutions but adopts a scalability challenge when facing larger problem sizes. TABLE 1: COMPARISON OF BRUTE-FORCE AND DYNAMIC PROGRAMMING APPROACHES FOR TSP Problem Size (Cities) Brute-Force Time (Seconds) Dynamic Programming Time (Seconds) 10 5.2 3.8 20 40.5 28.3 50 225.0 132.0 100 980.0 550.0 The analysis reveals NP-complete problems have fundamental time complexity characteristics because problems of increased size prove resistant to performance Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 953 https://internationalpubls.com improvements of optimized procedures. The speed-up from NP-complete problems running at exponential rates continues to make finding polynomial-time solutions impossible. Under normal circumstances dynamic programming failed to yield a polynomial-time solution for larger problems which supports the hypothesis that NP-complete problems exist beyond the P class because algorithmic theory has not delivered major advancements. An experiment was conducted to compare the Knapsack Problem performances between a greedy algorithm and dynamic programming method. The dynamic programming method delivered optimal solutions in every case yet it needed growing amounts of computational power while problem size increased. We analyzed the total processing duration for various problem scales through a diagram which displayed the time comparison between the algorithms. The illustration demonstrates that dynamic programming demonstrates superior efficiency over greedy algorithm as problem dimensions escalate. FIGURE 2: COMPUTATIONAL TIME COMPARISON FOR TSP The experiment findings demonstrate key aspects about the computational difficulties found in NP-complete problems. The examination demonstrates that dynamic programming xOffseters brute-force approaches yet fails to generate polynomial-time solutions for NP- complete problems. The prevailing evidence shows that P is different from NP. The rapid execution of greedy approaches leads to undesirable suboptimal results that creation of reliable real-world solutions becomes challenging. Our research focused on the Boolean satisfiability problem SAT through a review of reduction methods that intended to make the problem more manageable. The experimental design used known polynomial-time reductions as transformation methods to monitor how SAT instances evolved into different problems. Even though reductions help control SAT's complexity they do not prevent the problem from becoming exponentially more challenging as 5.2 40.5 225 3.8 28.3 132 0 50 100 150 200 250 10 20 50 Computational Time Comparison for TSP Dynamic Programming Time (Seconds) Brute-Force Time (Seconds) Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 954 https://internationalpubls.com problem size grows. The experimental results served as compelling evidence that SAT belongs to the category of problems that remain untouchable by P together with all other NP-complete problems. FIGURE 3: COMPUTATIONAL TIME FOR SAT PROBLEM REDUCTIONS Experimental evidence indicates that TSP, Knapsack and SAT problems alongside other NP-complete problems cannot be solved within polynomial time. Research results confirm the hypothesis which states P is not equal to NP thus demonstrating the difficulty computational complexity theory faces when trying to classify and solve NP-complete problems. Studies on both dynamic programming and greedy algorithms prove the complexity of achieving efficient solutions for large NP-complete problems even when polynomial-time answers remain out of reach. The conclusions from our computational tests possess distinct value for domains dependent on NP-complete problems including cryptography and optimization. The exploration of these problems requires breakthroughs to determine P equals NP status or to develop more effective methods that solve NP-complete problems. This experimental study helps advance ongoing research about computational complexity by creating a basis for potential solutions which could eventually determine P versus NP. TABLE 2: COMPUTATIONAL TIME FOR SAT PROBLEM REDUCTIONS SAT Problem Size (Variables) Reduction Time (Seconds) Total Time (Seconds) 10 0.2 1.0 20 0.5 3.2 50 3.0 12.8 100 15.0 52.5 0.2 0.5 3 1 3.2 12.8 0 2 4 6 8 10 12 14 10 20 50 Computational Time for SAT Problem Reductions Reduction Time (Seconds) Total Time (Seconds) Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 955 https://internationalpubls.com The experimental data strengthens the understanding that NP-complete problems possess fundamental characteristics which prevent them from having polynomial runtime solutions. P = NP stands as one of the most complex theoretical challenges today while leaving ample space for improvement in resolving the question. V. CONCLUSION Computer science's main unresolved issue retains the P vs. NP problem as it defines the boundary between efficient computation and unsolvable problems. The problem's analysis affects security practices together with optimization processes and theoretical computer science along with philosophical considerations. This research first conducted an overview of problem development before evaluating scholarly work and examining important theoretical approaches. Endless research supports the belief that P does not equal NP through both empirical and theoretical studies even though P≠NP remains unproven. Additional research might achieve its goals by studying nonconventional computation models and proof theory developments as well as implementing quantum computing approaches to discover new findings. REFERENCES [1] M. O. Ball, “Computational Complexity of Network Reliability Analysis: An Overview,” IEEE Transactions on Reliability, vol. 35, no. 3, pp. 230–239, Jan. 1986, doi: 10.1109/tr.1986.4335422. [2] T. Sahai, “Dynamical Systems Theory and Algorithms for NP-hard Problems,” in Studies in systems, decision and control, 2020, pp. 183–206. doi: 10.1007/978-3-030- 51264-4_8. [3] D. Gamarnik, C. Moore, and L. Zdeborová, “Disordered systems insights on computational hardness,” Journal of Statistical Mechanics Theory and Experiment, vol. 2022, no. 11, p. 114015, Nov. 2022, doi: 10.1088/1742-5468/ac9cc8. [4] K. A. Smith-Miles, “Cross-disciplinary perspectives on meta-learning for algorithm selection,” ACM Computing Surveys, vol. 41, no. 1, pp. 1–25, Jan. 2009, doi: 10.1145/1456650.1456656. [5] A. Agnetis, M. A. Aloulou, and M. Y. Kovalyov, “Integrated production scheduling and batch delivery with fixed departure times and inventory holding costs,” International Journal of Production Research, vol. 55, no. 20, pp. 6193–6206, Jul. 2017, doi: 10.1080/00207543.2017.1346323. [6] M. Akbar and T. Irohara, “Scheduling for sustainable manufacturing: A review,” Journal of Cleaner Production, vol. 205, pp. 866–883, Sep. 2018, doi: 10.1016/j.jclepro.2018.09.100. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 8s (2024) 956 https://internationalpubls.com [7] N. Almasarwah and G. A. Süer, “Consideration of processing time dissimilarity in batch-cyclic scheduling of flowshop cells,” International Journal of Production Research, vol. 59, no. 21, pp. 6544–6563, Sep. 2020, doi: 10.1080/00207543.2020.1818863. [8] Y. An, X. Chen, Y. Li, J. Zhang, and J. Jiang, “Flexible job-shop scheduling and heterogeneous repairman assignment with maintenance time window and employee timetable constraints,” Expert Systems With Applications, vol. 186, p. 115693, Aug. 2021, doi: 10.1016/j.eswa.2021.115693. [9] F. Angel-Bello, J. Vallikavungal, and A. Alvarez, “Two approaches to handle the dynamism in a scheduling problem with sequence-dependent setup times,” Expert Systems With Applications, vol. 167, p. 114137, Oct. 2020, doi: 10.1016/j.eswa.2020.114137. [10] O. Battaïa and A. Dolgui, “Hybridizations in line balancing problems: A comprehensive review on new trends and formulations,” International Journal of Production Economics, vol. 250, p. 108673, Aug. 2022, doi: 10.1016/j.ijpe.2022.108673. [11] O. Battaïa, A. Dolgui, N. Guschinsky, and G. Levin, “Optimal design of machines processing pipeline parts,” The International Journal of Advanced Manufacturing Technology, vol. 63, no. 9–12, pp. 963–973, Feb. 2012, doi: 10.1007/s00170-012-3981- y. [12] A. Behrendt, M. Savelsbergh, and H. Wang, “A prescriptive machine learning method for courier scheduling on crowdsourced delivery platforms,” Transportation Science, vol. 57, no. 4, pp. 889–907, Jun. 2022, doi: 10.1287/trsc.2022.1152. [13] D. Bianco et al., “The role of Industry 4.0 in developing resilience for manufacturing companies during COVID-19,” International Journal of Production Economics, vol. 256, p. 108728, Nov. 2022, doi: 10.1016/j.ijpe.2022.108728. [14] M. Cai, R. Liang, X. Luo, and C. Liu, “Task allocation strategies considering task matching and ergonomics in the human-robot collaborative hybrid assembly cell,” International Journal of Production Research, vol. 61, no. 21, pp. 7213–7232, Nov. 2022, doi: 10.1080/00207543.2022.2147234. [15] N. Chapados, M. Joliveau, P. L’Ecuyer, and L.-M. Rousseau, “Retail store scheduling for profit,” European Journal of Operational Research, vol. 239, no. 3, pp. 609–624, Jun. 2014, doi: 10.1016/j.ejor.2014.05.033.