Autonomous Manufacturing Systems Research Articles

Negotiation-based scheduling considering agent emotion

Scheduling in decentralized manufacturing systems is being expanded and transformed to a smart level where manufacturing tasks are executed autonomously. However, achieving self-organized manufacturing resources that can make collaborative decisions remains challenging. This paper presents a novel negotiation model for decentralized scheduling in an autonomous manufacturing system. A variety of emotions represent the willingness of autonomous intelligent agents (IAs) to negotiate for a consensus despite their diverse goals. The proposed communication protocol considers the emotional aspects of the autonomous machine agents in decentralized decision-making. The negotiation model focuses on analyzing the effects of two key emotions: selfishness and cooperation. This study constitutes the basis for experiments on multi-goal machine tasks and artificial emotion mechanism in a flow shop production system. The complexities of negotiation dynamics between IAs, where the interplay of selfish and cooperative strategies influences the outcomes and efficiency of reaching agreements, are explored. The experimental results show that selfish IAs eventually reach agreements through gradual compromise, but cooperative IAs, with their flexibility, can quickly achieve resolutions, especially when bargaining power is balanced. Aligning with the dual concern theory, this study reveals that while selfish IAs can dominate cooperative ones, the cooperative approach often leads to faster and more effective outcomes, regardless of the level of selfishness involved.

G-code evaluation in CNC milling to predict energy consumption through Machine Learning

Computerized Numeric Control (CNC) plays an essential role in highly autonomous manufacturing systems for interlinked process chains for machine tools. NC-programs are mostly written in standardized G-code. Evaluating CNC-controlled manufacturing processes before their real application is advantageous due to resource efficiency. One dimension is the estimation of the energy demand of a part manufactured by an NC-program, e.g. to discover optimization potentials. In this context, this paper presents a Machine Learning (ML) approach to assess G-code for CNC-milling processes from the perspective of the energy demand of basic G-commands. We propose Latin Hypercube Sampling as an efficient method of Design of Experiments to train the ML model with minimum experimental effort to avoid costly setup and implementation time of the model training and deployment.

An Additive Manufacturing Testbed to Evaluate Machine Learning-Based Autonomous Manufacturing

Abstract This paper details the design and operation of a testbed to evaluate the concept of autonomous manufacturing to achieve a desired manufactured part performance specification. This testbed, the autonomous manufacturing system for phononic crystals (AMSPnC), is composed of additive manufacturing, material transport, ultrasonic testing, and cognition subsystems. Critically, the AMSPnC exhibits common manufacturing deficiencies such as process operating window limits, process uncertainty, and probabilistic failure. A case study illustrates the AMSPnC function using a standard supervised learning model trained by printing and testing an array of 48 unique designs that span the allowable design space. Using this model, three separate performance specifications are defined and an optimization algorithm is applied to autonomously select three corresponding design sets to achieve the specified performance. Validation manufacturing and testing confirms that two of the three optimal designs, as defined by an objective function, achieve the desired performance, with the third being outside the design window in which a distinct bandpass is achieved in phononic crystals (PnCs). Furthermore, across all samples, there is a marked difference between the observed bandpass characteristics and predictions from finite elements method computation, highlighting the importance of autonomous manufacturing for complex manufacturing objectives.

ManuChain II: Blockchained Smart Contract System as the Digital Twin of Decentralized Autonomous Manufacturing Toward Resilience in Industry 5.0

In Industry 5.0 vision, machines are empowered with the interaction capability to autonomously make local decisions and coordinate with each other as well as humans. However, how to form a group consensus on the rapid self-organizing of the manufacturing process is critical for achieving manufacturing resilience under disturbances and disruptions. Based on our formerly developed system ManuChain (Leng et al., 2020), this article proposes a blockchained smart contract system (BSCS), named ManuChain II, as the digital twin of a decentralized autonomous manufacturing system for achieving resilience in Industry 5.0. A blockchain-secured multiagent system architecture together with a product data model is established to form the BSCS. The BSCS could prevent tampering with data and enhance the transparency of the product manufacturing process. In BSCS, two types of blockchained smart contracts (SCs) are established with a bi-level interplay computing architecture. The lower-level contracts perform the predefined and learned patterns of task coordination for achieving resilience under internal disruptions. The upper-level contracts are incorporated with communication-efficient decentralized deep-learning algorithms for the learning, updating, transferring, and sharing of coordination patterns used in the autonomous decision in lower-level SCs, thereby achieving the continuous improvement of the system’s decentralized autonomous intelligence. Via incorporating the decentralized deep learning algorithms into blockchained SCs, this study reveals a new way to realize the self-organizing intelligence of the manufacturing system for enhancing resilience toward Industry 5.0.

A maturity model for the autonomy of manufacturing systems

Modern manufacturing has to cope with dynamic and changing circumstances. Market fluctuations, the effects caused by unpredictable material shortages, highly variable product demand, and worker availability all require system robustness, flexibility, and resilience. To adapt to these new requirements, manufacturers should consider investigating, investing in, and implementing system autonomy. Autonomy is being adopted in multiple industrial contexts, but divergences arise when formalizing the concept of autonomous systems. To develop an implementation of autonomous manufacturing systems, it is essential to specify what autonomy means, how autonomous manufacturing systems are different from other autonomous systems, and how autonomous manufacturing systems are identified and achieved through the main features and enabling technologies. With a comprehensive literature review, this paper provides a definition of autonomy in the manufacturing context, infers the features of autonomy from different engineering domains, and presents a five-level model of autonomy — associated with maturity levels for the features — to ensure the complete identification and evaluation of autonomous manufacturing systems. The paper also presents the evaluation of a real autonomous system that serves as a use-case and a validation of the model.

Reinforcement Learning Enabled Autonomous Manufacturing Using Transfer Learning and Probabilistic Reward Modeling

Here we propose a reinforcement learning enabled physical autonomous manufacturing system (AMS) that is capable of learning the manufacturing process parameters to autonomously fabricate a complex-geometry artifact with desired performance characteristics. The poor sample efficiency of traditional RL algorithms challenges real-world manufacturing decision making due to a high variable cost from raw material, machine utilization, and labor costs. To make decision making sample efficient, we propose to leverage a first-principles based source task for training, transfer effective representations from trained knowledge, and then use these representations to interact with the physical system to learn a probabilistic model of the target reward function. We deploy this idea to a novel dataset obtained from a custom physical AMS machine that can autonomously manufacture phononic crystals, a complex geometry artifact with spectral response as performance characteristic. We demonstrate that our method uses as low as 25 artifacts to model the interesting part of the target reward function and find an artifact with high reward. This task typically requires manual design of phononic crystals and extensive empirical iterations on the order of hundreds.

Digital Twin Modeling in Virtual Enterprises and Autonomous Manufacturing Systems: Deep Learning and Neural Network Algorithms, Immersive Visualization Tools, and Cognitive Data Fusion Techniques

Digital Twin Modeling in Virtual Enterprises and Autonomous Manufacturing Systems: Deep Learning and Neural Network Algorithms, Immersive Visualization Tools, and Cognitive Data Fusion Techniques

Complexity theory and self-organization in Cyber-Physical Production Systems

The heterogeneity of the components of a Cyber-Physical Production System in addition to the high decentralization and autonomy required in Industry 4.0, introduces new levels of engineering complexity and dynamism that classical reductionists approaches are not able to solve. Within this context, novel solutions that rely on complexity sciences seem to be a good alternative to cope with these underline challenges. In this context, this paper presents a conceptual framework of complexity theory, self-organization and emergence and its subsequent relation to cyber manufacturing systems. Such analysis shows very promising ideas in the further development of complex, robust, adaptive and at least partial autonomous manufacturing systems.

ResourceNet: a collaboration network among decentralised manufacturing resources for autonomous exception‐handling in smart manufacturing

The fast-changing market and increasing demand for customised products have imposed manufacturers to improve the flexibility and robustness of their manufacturing execution systems. The ability to recover from exceptional events is fundamental to autonomous manufacturing systems in the era of smart manufacturing. Currently, the various types of manufacturing exceptions remain unclear. This research investigated the typical exceptions and proposed a general framework that supports the autonomous exception-handling and resource discovery in dynamic environments. A multi-layer peer-to-peer network is used to model the resources, services, and events in decentralised manufacturing systems. The exception-handling mechanism is designed that incorporates rule-based reactions, matching of features, recording of behaviour patterns etc. The feasibility of the proposed methods is also discussed, which shows many exceptions that were ignored in previous scheduling models can be timely identified and easily handled. This research explores the complex relations among manufacturing resources and provides an intelligent overall framework for self-organising manufacturing with self-diagnose capabilities.

Task scheduling system for UAV operations in agricultural plant protection environment

Unmanned aerial vehicle (UAV) can greatly reduce manpower in agricultural plant protection such as watering, sowing, pesticide spraying. It helps in creating an autonomous manufacturing system by executing tasks with less human intervention in time-efficient manner. Consequently, reasonable and efficient planning is one essential component to be focused on; yet to the best of our knowledge, there are few research on practical agricultural plant protection UAV scheduling. This work proposes method to solve the Agricultural Plant Protection UAV scheduling problem in practice such as pesticide spraying, flying and charging. At first, tasks are assigned to UAVs, then schedules are planned for UAVs. To find a near optimal schedule quickly, our method is incorporated with Dragonfly Algorithm. This proposed method is implemented and tested on datasets generated based on a real agricultural plant protection environment. Performance evaluation of our method is discussed in detail and the set of parameters that determining the best solution is reported.

Robots assembling machines: learning from the World Robot Summit 2018 Assembly Challenge

ABSTRACTThe Industrial Assembly Challenge at the World Robot Summit was held in 2018 to showcase the state-of-the-art of autonomous manufacturing systems. The challenge included various tasks, such as bin picking, kitting, and assembly of standard industrial parts into 2D and 3D assemblies. Some of the tasks were only revealed at the competition itself, representing the challenge of ‘level 5’ automation, i.e. programming and setting up an autonomous assembly system in less than one day. We conducted a survey among the teams that participated in the challenge and investigated aspects such as team composition, development costs, system setups as well as the teams' strategies and approaches. An analysis of the survey results reveals that the competitors have been in two camps: those constructing conventional robotic work cells with off-the-shelf tools, and teams who mostly relied on custom-made end effectors and novel software approaches in combination with collaborative robots. While both camps performed reasonably well, the winning team chose a middle ground in between, combining the efficiency of established play-back programming with the autonomy gained by CAD-based object detection and force control for assembly operations.

The importance of prediction methods in industry 4.0 on the example of steel industry

Abstract This paper presents the importance of the prediction of steel production in industry 4.0 along with forecasts for steel production in the world until 2022. In the last two decades, the virtual world has been increasingly entering production. Today’s manufacturing systems are becoming faster and more flexible – easily adaptable to new products. Steel is the basic structural material (base material) for many industrial sectors. Industries such as automotive, mechanical engineering, construction and transport use steel in their production processes. Prediction methods in cyber-physical production systems are gaining in importance. The task of prediction is to reduce risk in the decision-making process. In autonomous manufacturing systems in industry 4.0 the role of prediction is more active than passive. Forecasts have the following functions: warning, reaction, prevention, normative, etc. The growing number of customized solutions in industry 4.0 translates into new challenges in the production process. Manufacturers must respond to individual customer needs more quickly, be able to personalize products while reducing energy and resource costs (saving energy and resources can increase the product competitiveness). The modern market becomes increasingly unpredictable. Production prediction under such conditions should be carried out continuously, which is possible because there is more empirical data and access to data. Information from the ongoing monitoring of the company’s production is directly transferred to the prospective evaluation. In view of the contemporary reciprocal use of automation, data processing, data exchange and manufacturing techniques, there is greater access to external data, e.g. on production in different target markets and with global, international, national, regional coverage. Companies can forecast in real time, and the forecasts obtained give the possibility to quickly change their production. Industry 4.0 (from the business objective point of view) aims to provide companies with concrete economic benefits – primarily by reducing manufacturing costs, standardizing and stabilizing quality, increasing productivity. Industry 4.0 aims to create a given autonomous smart factory system in which machines, factory components and services communicate and cooperate with each other, producing a personalized product. The aim of this paper is to present new challenges in the production processes in relation to steel production, as well as to prepare and present forecasts of (quantitative) steel production of territorial, global and temporary range until 2022, taking into account the applied production technologies (BOF and EAF). For forecasting purposes, classic trend models and adaptive trend models were used. This methodology was used to build separate forecasts for: total steel production, BOF steel and EAF steel. Empirical data is world steel production in 2000-2017 (annual production volume in Mt).

Triple state reliability measurement for a complex autonomous robot system based on extended triangular distribution

Modern manufacturing systems should comply with high reliability and availability standards in order to guarantee efficient planning and optimization. Manufacturing systems including autonomous robots make the problem more challenging. By increasing the number of components in the autonomous manufacturing system and with respect to their intermittent interactions, a complex system is evolved. The reliability of such a system is significant due to high repair costs and failure downtime duration. Therefore, in this paper a reliability measurement framework is developed for a triple state autonomous robot system in an automated manufacturing environment. In this framework, an extension of the triangular probability distribution is proposed. In an implementation study, the effectiveness and computational efficiency of the proposed method is illustrated. In addition, an analysis on the failure rate using the maximum likelihood estimation is reported.

Attribute-based identification processes for autonomous manufacturing systems – an approach for the integration in factory planning methods

The demand for customer innovated products is rising. In order to deal with the resulting complexity, the factory planner can employ autonomous manufacturing systems, in which objects are enabled to coordinate themselves through production systems. A fundamental requirement for this kind of control is the identification of objects. In this approach, attribute-based identification is focused on decreasing the identification costs. Instead of using an ID, an object type is recognised by its attributes. The integration of such a system, especially the required sensors, in an existing plant is difficult. Therefore, planning and development should take place within the factory planning.

Characterization of Autonomous Production by a Stage Model

Manufacturing systems aim to optimize criteria such as resource efficiency, profitability and flexibility. Typically, these criteria form a conflict of interests. Driven by recent trends like the Digital and Biological Transformation, manufacturing systems try to overcome this conflict by heading towards autonomy. For the realization of an autonomous manufacturing system, stages of development need to be identified. Therefore, this paper evaluates different concepts for manufacturing systems (e.g. Fractal Factory, Holonic Factory, etc.) to derive characteristics for a stepwise development towards Autonomous Production. As a result, a target state and a stage model for Autonomous Production are defined.

Analysis of Smart Manufacturing Technologies for Industry Using AI Methods

Smart manufacturing technologies have gained significant attention in the industrial sector due to their potential to revolutionize traditional manufacturing processes. Among these technologies, artificial intelligence (AI) methods have emerged as powerful tools for enhancing efficiency, productivity, and decision-making in manufacturing operations. This paper presents an analysis of smart manufacturing technologies for industry using AI methods. The analysis focuses on the application of AI techniques such as machine learning, deep learning, and data analytics in various aspects of smart manufacturing, including predictive maintenance, process optimization, quality control, and supply chain management. The paper provides an overview of the key AI methods employed in smart manufacturing and discusses their benefits and challenges. It also highlights case studies and real-world implementations of AI-based smart manufacturing systems. The findings of this analysis demonstrate the significant contributions of AI methods in enabling intelligent and autonomous manufacturing systems. The paper concludes with insights into the future directions and potential impact of AI-driven smart manufacturing technologies in industry, emphasizing the importance of continued research and development in this field to unlock the full potential of smart manufacturing in the industry. Smart manufacturing technologies have revolutionized the industrial sector by enhancing productivity, efficiency, and flexibility. Artificial intelligence (AI) methods, such as machine learning and data analytics, play a crucial role in enabling smart manufacturing systems to optimize processes and make informed decisions. This research paper aims to analyse the application of AI methods in smart manufacturing technologies. The study explores various AI-based approaches used in different stages of smart manufacturing, including data acquisition, data analysis, process optimization, and predictive maintenance. The research provides insights into the benefits, challenges, and potential future developments of AI in smart manufacturing, offering valuable guidance for industries aiming to implement these technologies.

Task scheduling system for UAV operations in indoor environment

The application of unmanned aerial vehicle (UAV) in indoor environment is emerging nowadays due to the advancements in technology. UAV brings more space flexibility in an occupied or hardly accessible indoor environment, e.g. shop floor of manufacturing industry, greenhouse, and nuclear powerplant. UAV helps in creating an autonomous manufacturing system by executing tasks with less human intervention in a time-efficient manner. Consequently, a scheduler is an essential component to be focused on; yet the number of reported studies on UAV scheduling has been minimal. This work proposes a mathematical model of the problem and a heuristic-based methodology to solve it. To suit near real-time operations, a quick response towards uncertain events and a quick creation of new high-quality feasible schedule are needed. Hence, the proposed heuristic is incorporated with particle swarm optimization algorithm to find a near optimal schedule in a short computation time. This proposed methodology is implemented into a scheduler and tested on a few scales of datasets generated based on real flight demonstrations. Performance evaluation of scheduler is discussed in detail, and the best solution obtained from a selected set of parameters is reported.

Innovations in digital modelling for next generation manufacturing system design

Interlinked and autonomous manufacturing systems provide new opportunities in smart manufacturing. Today's manufacturing system design processes and architecture still are based on traditional engineering methods and can hardly cope with increased system complexity. Hence new cyber physical production systems (CPPS) design and architecture principles as well as corresponding validation and verification methods are necessary. This paper presents firstly a new architecture design approach for modularised design of CPPS. Secondly, new capabilities in designing with modular construction kits, simulating functional behaviour and validating with virtual, functional prototype are introduced. Thirdly, a proposal for the development approach and virtual validation is presented and risks as well as challenges are discussed.

Recent Advances in Control Theory and Applications in Turkey

TOK’14, the National Symposium on Automatic Control, is an annually organized event bringing together the researchers from various universities and industry in Turkey. The symposium in 2014 was hosted by the Kocaeli University on 11–13 September in Kocaeli. The current special issue contains eight papers from the symposium programme, which had more than 200 contributions. Control theory is used for strategically cutting-edge high technologies, which provide important support for the sustainable development of flexible and autonomous manufacturing, medicine, automative, space, aerospace, finance, robotics, power and energy systems, defence, networking, etc. The focus of the special issue is to report the recent advances in control theory and applications in Turkey. Recently, national conferences and symposia are becoming hosts for more qualified research studies and industrial projects, as the research and development funds supported by the Turkish government have been rapidly increased. The goal of this special issue is to provide a forum for the latest research and projects, focusing on control theory and their industrial applications. The first paper, contributed by Morgul et al., proposed an alternative approach that steers away from explicit mechanical modelling towards data-driven system identification employed in vertical hopping robots. This identification approach treats the system as a black box and produces empirical models using harmonic transfer functions (HTF). Using this method reduces the complexity of the mathematical models and approximates unmodelled dynamics in legged locomotion robots. They showed that HTFs can be used as predictors of simple locomotion models on training data and their performance on sinusoid inputs is much better than on step inputs. The second paper, by Ozdogan and Leblebicioglu, presented modelling of a single axis (inner azimuth gimbal) of a four-axis gyro-stabilized electro-optic gimbal system using frequency response function (FRF). FRF is existing technique in the literature but its application to an electro-optic gimbal system is a challenging problem. The third paper, contributed by Efe et al., compared the performance of the controllers, namely proportional–integral–derivative control, sliding mode control, backstepping control, feedback linearization-based control and fuzzy control, based on integral absolute error (IAE), maximum absolute error (MAE), integral squared error (ISE), integral time squared error (ITSE) and error variance (EV) for Quadrotor-type aerial vehicles. They found that sliding mode control is better than the others in terms of simplicity and closed-loop control performance. In the fourth paper, Iplikci and Bahtiyar developed the Runge–Kutta model predictive controller (RKMPC) implemented in real time on a field-programmable gate array (FPGA) in an electromagnetic levitation system. They compared performance of the RKMPC with a conventional nonlinear model predictive controller (NMPC) method for step and sinusoidal inputs. They found that the RKMPC provides very satisfactory control performance for a very fast experimental non-linear system. The fifth paper, authored by Alagoz and Kaygusuz, examined smart grid energy market management performance of a fractional-order proportional–integral (PI) controller in the case of communication and operation delays in a multisource energy market model. The simulation results showed that using of the fractional-order PI controller decreases the average energy shortage error and price volatility in the smart grid energy markets. Hence, it can be a good candidate for automated market management applications. In the sixth paper, Bagis and Konar studied a performance comparison of the Sugeno and Mamdani-type fuzzy models optimized by the artificial bee colony (ABC) algorithm with the other modelling approaches in the literature for a microstrip antenna. The ABC algorithm is found to be a powerful candidate for solving the numerical optimization of fuzzy models. In the seventh paper, Leblebicioglu et al. solved the path planning problem using a genetic algorithm (GA) with the proposal of novel evolutionary operators for multiple unmanned aerial vehicles (UAVs). They obtained nearly global optimum flight path solutions regarding UAV dynamics. Real-world experiments conducted using small UAVs with

Adaptive process control based on a self-learning mechanism in autonomous manufacturing systems

To survive in the highly competitive global economy, manufacturing systems must be able to adapt to new circumstances. An important prerequisite for adaptation is the ability to learn, a process based on knowledge discovery and growth. The aim of this research is to uncover knowledge by examining a large volume of real-time manufacturing data collected during manufacturing operations and to use the insights gained to support decision-making and adaptive process control. The paper presents the concept of a self-learning autonomous work system. This concept introduces a learning loop into a manufacturing system composed of data acquisition, data mining (DM), and knowledge-building models. Two methods for DM are applied. A descriptive DM method enables discovery of patterns in data that may contribute to a better understanding of the manufacturing processes. A predictive process provides knowledge in the form of rules, which can then be used for enhanced decision-making. To illustrate the utility of the knowledge models, the concept of adaptive process control is introduced and implemented in a high pressure die-casting domain. A case study based on industrial data collected during die-casting operations provides a demonstration of the concept.