Industrial Use-case Research Articles

Upper-Limb and Low-Back Load Analysis in Workers Performing an Actual Industrial Use-Case with and without a Dual-Arm Collaborative Robot

Upper-Limb and Low-Back Load Analysis in Workers Performing an Actual Industrial Use-Case with and without a Dual-Arm Collaborative Robot

Increasing the Reliability of Software Systems Using a Large-Language-Model-Based Solution for Onboarding

Software systems are often maintained by a group of experienced software developers in order to ensure that faults that may bring the system down are less likely. Large turnover in organizations such as CERN makes it important to think of ways of onboarding newcomers on a technical project rapidly. This paper focuses on optimizing the way that people get up-to-speed on the business logic and technologies used on the project by using a knowledge-imbued large language model that is enhanced using domain-specific knowledge from the group or team’s internal documentation. The novelty of this approach is the gathering of all of these different open-source methods for developing a chatbot and using it in an industrial use-case.

A Lean approach in the upstream phase of the product life cycle: Lean Enterprise Model practices applications and analysis from industrial use-cases

A Lean approach in the upstream phase of the product life cycle: Lean Enterprise Model practices applications and analysis from industrial use-cases

Blockchain-Based Smart Contract and Edge AI Applied in a Multirobot System: An Approach

Mining exploration is a common use case in Industry 4.0, which requires robots to work collaboratively to achieve desired goals. In this context, creating and orchestrating software for robotics is often a difficult task.

Stakeholder-driven Design and Exploration of Sociotechnical Systems using Virtual-Reality

With ever-shorter product and technology lifecycles, companies are pressured to shorten the time-to-market for new products. This requires faster and more agile design processes both in product development and in setting up new production environments. Technological advances in the field of immersive technologies, such as extended-reality, provide tools to address these challenges and increase process agility. One tool is virtual prototyping. Virtual prototyping enables the involvement of users and other stakeholders early in the development process. Currently, the use of virtual prototypes is largely limited to design reviews and assessments of existing system concepts. Inspired by the paradigms of creative collaboration (co-creation) and exploration, we have developed a participatory design method where users and other stakeholders can explore and directly adapt virtual prototypes within the virtual environment to find the optimal design concept. In this paper, we present an exploratory and stakeholder-driven design paradigm and its application in the context of two industrial use-cases in product development and production training. We illustrate how virtual exploration environments can be built, how the exploration and design process is orchestrated with the involvement of all relevant stakeholders, and how the results and findings can be utilized in later system development phases.

An Industrial Use-Case for Reinforcement Learning: Optimizing a Production Planning in Stochastic Conditions

An Industrial Use-Case for Reinforcement Learning: Optimizing a Production Planning in Stochastic Conditions

PAS - A Feature Selection Process Definition for Industrial Settings

Industry 4.0, and the accompanying digitalization of industrial processes, lead to a continuously increasing amount of data. One reason is the installation of more sensors on machines and production lines in combination with shorter measurement intervals. The rising amount of data requires elaborate methods to remove redundant and uninformative data to reduce time complexity and required computational resources. This could be achieved by feature selection methods in combination with the right preprocessing and enhancement methods to generate and identify important features. Feature selection methods are mostly described in a methodological way without knowing how data is handled prior to their application, assuming a high-quality dataset. To fill this gap, we define a process for feature selection, focusing on time-series industrial data. The definition consists of three phases for the data adaptation: Preprocessing, Amplification and Selection. Each phase fulfills a specific purpose, ranging from handling missing and censored data to the enhancement of the dataset by higher-order data and aggregations and the selection itself. Our experiments show that the application of these phases on three different industrial use-cases lead to better accuracy results in machine learning models.

Collision detection for collaborative assembly operations on high-payload robots

Human-robot collaboration for high-payload industrial applications has the potential to unlock new applications and improve operator ergonomics. However, ensuring safety with closer proximity remains challenging due to the large payload. Collision detection can improve safety, but must be modified to consider the dynamics of high payload applications (internal oscillation, effective low-pass filtering of external force), while the higher inertia raises questions about the resulting collision forces. This paper proposes the first collision detection approach for high-payload applications, using a force/torque sensor at the end-effector and motor current measurements for redundancy to be compliant with ISO/TS 15066:2016. Analysis of a linear model informs the design of the detection algorithm and effect of higher payload. Experimental validation shows feasibility and performance of collision detection over a range of motion and payload conditions, and the impact of the safety system in improving an industrial use-case.

Preliminary results of a biomechanical risk assessment of an actual industrial use-case executed with and without a dual-arm CoBot

Preliminary results of a biomechanical risk assessment of an actual industrial use-case executed with and without a dual-arm CoBot

Multilevel Subgranting by Power of Attorney and OAuth Authorization Server in Cyber–Physical Systems

Many Cyber-Physical Systems are today semiautonomous and powerful enough to perform advanced tasks on their own. This means they can also act as representatives of people or devices that have given them an order. However, traditional access control policies and delegation models do not meet industrial requirements such as support for letting autonomous CPS devices act on their own with certified credentials under the sub authorization by subcontractors, without the need for a separate account per device. In this paper, we analyze and compare power of attorney, proxy signature by warrant, and OAuth to identify the strengths and challenges of each. Based on the comparison, we propose an OAuth grant type based on the power of attorney and inspired by the concept of proxy signature by warrant. Power of Attorney is a generic and self-contained document that a principal signs and directs to an agent, thereby providing it the power to execute actions on behalf of the principal for a predefined time, even if it is offline. One key advantage of the power of attorney is that it can support effective sub-granting on several levels to support industrial scenarios where resource owners bring in authorized contractors that can in their turn authorize and bring in several devices without incurring management overhead to the resource owner. A proof-of-concept and performance evaluation of the proposed model is presented using an industrial use-case scenario with multi-level authorization.

Non-Fungible Token based Smart Manufacturing to scale Industry 4.0 by using Augmented Reality, Deep Learning and Industrial Internet of Things

The recent revolution in Industry 4.0 (IR 4.0) has characterized the integration of advance technologies to bring the fourth industrial revolution to scale the manufacturing landscape. There are different key drivers for this revolution, in this research we have explored the following among them such as, Industrial Internet of Things (IIoT), Deep Learning, Blockchain and Augmented Reality. The emerging concept from blockchain namely “Non-Fungible Token” (NFT) relating to the uniqueness of digital assets has vast potential to be considered for physical assets identification and authentication in the IR 4.0 scenario. Similarly, the data acquired through the deployment of IIoT devices and sensors into smart industry spectrum can be transformed to generated robust analytics for different industry use-cases. The predictive maintenance is a major scenario in which early equipment failure detection using deep learning model on acquired data from IIoT devices has major potential for it. Similarly, the augmented reality can be able to provide real-time visualization within the factory environment to gather real-time insight and analytics from the physical equipment for different purposes. This research initially conducted a survey to analyse the existing developments in these domains of technologies to further widen its horizon for this research. This research developed and deployed a smart contract into an ethereum blockchain environment to simulate the use-case for NFT for physical assets and processes synchronization. The next phase was deploying deep learning algorithms on a dataset having data generated from IIoT devices and sensors. The Feedforward and Convolutional Neural Network were used to classify the target variables in relation with predictive maintenance failure analysis. Lastly, the research also proposed an AR based framework for the visualization ecosystem within the industry environment to effectively visualize and monitory IIoT based equipment’s for different industrial use-cases i.e., monitoring, inspection, quality assurance.

Bringing Energy Efficiency Closer to Application Developers: An Extensible Software Analysis Framework

Green, sustainable and energy-aware computing terms are gaining more and more attention during the last years. The increasing complexity of Internet of Things (IoT) applications makes energy efficiency an important requirement, imposing new challenges to software developers. Software tools capable of providing energy consumption estimations and identifying optimization opportunities are critical during all the phases of application development. This work proposes a novel framework that targets the energy efficiency at application development level. The proposed framework is implemented as a single user-friendly tool-flow, providing a variety of useful features, such as the estimation of the energy consumption without the need of executing the application on the targeted IoT devices and the estimation of potential gains by GPU acceleration on modern heterogeneous IoT architectures. The proposed methodology provides several novel contributions, such as the combination of static analysis and dynamic instrumentation approaches in order to exploit the advantages of both. The framework is evaluated on widely used benchmarks, achieving increased estimation accuracy (more than 90% for similar architectures and more than 72% for the potential use of the GPU). The effectiveness of the framework is further demonstrated using two industrial use-cases achieving an energy reduction from 91% up to 98%.

A novel approach for automatic annotation of human actions in 3D point clouds for flexible collaborative tasks with industrial robots.

Manual annotation for human action recognition with content semantics using 3D Point Cloud (3D-PC) in industrial environments consumes a lot of time and resources. This work aims to recognize, analyze, and model human actions to develop a framework for automatically extracting content semantics. Main Contributions of this work: 1. design a multi-layer structure of various DNN classifiers to detect and extract humans and dynamic objects using 3D-PC preciously, 2. empirical experiments with over 10 subjects for collecting datasets of human actions and activities in one industrial setting, 3. development of an intuitive GUI to verify human actions and its interaction activities with the environment, 4. design and implement a methodology for automatic sequence matching of human actions in 3D-PC. All these procedures are merged in the proposed framework and evaluated in one industrial Use-Case with flexible patch sizes. Comparing the new approach with standard methods has shown that the annotation process can be accelerated by 5.2 times through automation.

Machine Learning in 6G Wireless Communications

Mobile communication systems are not only the core of the Information and Communication Technology (ICT) infrastructure but also that of our social infrastructure. The 5th generation mobile communication system (5G) has already started and is in use. 5G is expected for various use cases in industry and society. Thus, many companies and research institutes are now trying to improve the performance of 5G, that is, 5G Enhancement and the next generation of mobile communication systems (Beyond 5G (6G)). 6G is expected to meet various highly demanding requirements even compared with 5G, such as extremely high data rate, extremely large coverage, extremely low latency, extremely low energy, extremely high reliability, extreme massive connectivity, and so on. Artificial intelligence (AI) and machine learning (ML), AI/ML, will have more important roles than ever in 6G wireless communications with the above extreme high requirements for a diversity of applications, including new combinations of the requirements for new use cases. We can say that AI/ML will be essential for 6G wireless communications. This paper introduces some ML techniques and applications in 6G wireless communications, mainly focusing on the physical layer.

Ultra Wide Band communication for condition-based monitoring, a bridge between edge and cloud computing

Ultra Wide Band (UWB) wireless communications offer a radically different approach to wireless communication compared to conventional narrow band systems in an industrial context. A lot of industrial applications do not require real-time acquisition and uses low data rates, but some scenarios, such as the condition monitoring systems and the control processes require higher data rates (i.e. tens of kHz). General purpose wireless solutions do not compliant industrial standards. This paper discusses the potential of UWB technology for realizing use cases in industry 4.0 and shows the implementation of an UWB sensor network for machine vibration monitoring that is able to acquire and transmit (6.8Mbit/s data rate) accelerometric measures sampled up to 32KHz with a maximum Packet Loss Rate (PLR) of 2.44%. It also shows as UWB could be a valid instrument to connect cloud computing and edge computing even for real time extraction of diagnostic and prognostic features from high-frequency signals. Different experiments have been performed in an industrial and laboratory environment to evaluate the performance of UWB radio for communication in a wireless sensor network.

Gathering Expert Knowledge in Process Industry

Knowledge intensive processes are often driven and constrained by mental models of experts. Especially in industrial processes, the knowledge is not explicit but tacit in many cases, which makes it difficult to provide verbal descriptions. Modeling and prediction tasks are often not compatible or differ from the view of experts if no additional knowledge regarding the process itself or important parts of the process are identified and integrated. In industrial environments, the required knowledge is present in operators, which have to deal with the task or process on a frequent basis. Including operators during the development process may increase the quality and performance as well as trust in the resulting application. Data scientists face great difficulties in eliciting and representing the knowledge of the experts in these systems for solving the given tasks. In this paper, we will provide a review of strategies and approaches for knowledge elicitation in industrial environments. The approaches focus on applicability in industrial use-cases, whereat the operators are the main source for knowledge. The differences of the approaches are highlighted and use-cases for their application are presented. Furthermore, we will provide a discussion on using these approaches at different stages during a project as usually multiple knowledge elicitation phases are required to reach a certain goal.

Safe Reinforcement Learning for Autonomous Navigation of a Driveable Vertical Mast Lift

In this work, we consider the issue of controlling a Driveable Vertical Mast Lift (DVML) to autonomously navigate an environment while ensuring required safety constraints. DMVL are industrial vehicles used in several applications, f.i. in logistics and smart agriculture, to allow operators placed in a basket accessing elevated worksites. When driving such machines, operators are exposed to hazards that could lead to potentially serious accidents. Reinforcement Learning (RL) is a data-driven approach that is increasingly being used to control complex systems. This work investigates the advancements in the field of Safe RL from a practical perspective, employing several state-of-art algorithms to equip a DVML with autonomous driving capabilities. We highlight how benchmark environments, while satisfactorily affirming Safe RL methodologies as proof-of-concepts, can widen the gap that prevents such methodologies from both being applied in real scenarios and becoming much more popular in industrial use-cases.

Heterogeneous Ensemble Model to Optimize Software Effort Estimation Accuracy

The software industry has rapidly expanded, and software development is now essential to the success of many multinational corporations. The demand for complex software systems has also dramatically increased. Effective software development is essential, given the limitations of resources such as money, time, and labor. The cost and effort calculations affect the development process and client needs. Project failure was usually caused by errors in job estimating. Underestimating could have severe repercussions, if the project’s budget ends up being more than anticipated. Project overruns, on the other hand, can also have a detrimental impact on how successfully software projects are finished. Researchers and experts in the field of software are investigating ways to keep management and development productivity at high levels. However, a stand-alone estimation has not produced any noteworthy research results. Over the last decade, the standalone estimating models have revealed inadequacies. Literature suggests opting ensemble models, would yield better results as compared to standalone models. We have proposed a heterogeneous ensemble effort estimation (EEE) model in this research. The proposed model is comprises of standalone estimating models such as 1) Use Case Point, 2) Expert Judgment (EJ) and 3)Artificial neural network (ANN). Finally, using linear combination rules, the effort of each unique base model is combined. We applied our ensemble model on the benchmark dataset i.e. ISBSG using three different variations to avoid biasness. Furthermore, trained models were applied on industry use-cases for cross-validation. The findings of this study demonstrated that, in comparison to stand-alone estimate strategies, the ensemble technique gave better estimation results.

Detecting Anomalous Multivariate Time-Series via Hybrid Machine Learning

This article investigates the use of hybrid machine learning (HML) for the detection of anomalous multivariate time-series (MVTS). Focusing on a specific industrial use-case from geotechnical engineering, where hundreds of MVTS need to be analyzed and classified, has permitted extensive testing of the proposed methods with real measurement data. The novel hybrid anomaly detector combines two means for detection, creating redundancy and reducing the risk of missing defective elements in a safety relevant application. The two parts are: 1) anomaly detection based on approximately 50 physics-motivated key performance indicators (KPIs) and 2) an unsupervised variational autoencoder (VAE) with long short-term memory layers. The KPI captures expert knowledge on the properties of the data that infer the quality of produced elements; these are used as a type of auto-labeling. The goal of the extension using machine learning (ML) is to detect anomalies that the experts may not have foreseen. In contrast to anomaly detection in streaming data, where the goal is to locate an anomaly, each MVTS is complete in itself at the time of evaluation and is categorized as anomalous or nonanomalous. The article compares the performance of different VAE architectures [e.g., long short-term memory (LSTM-VAE) and bidirectional LSTM (BiLSTM-VAE)]. The results of using a genetic algorithm to optimize the hyperparameters of the different architectures are also presented. It is shown that modeling the industrial process as an assemblage of subprocesses yields a better discriminating power and permits the identification of interdependencies between the subprocesses. Interestingly, different autoencoder architectures may be optimal for different subprocesses; here two different architectures are combined to achieve superior performance. Extensive results are presented based on a very large set of real-time measurement data.

Stochastic Game for Resource Management in Cellular Zero-Touch Deterministic Industrial M2M Networks

In this letter, we investigate the resource allocation problem for cellular Zero-Touch Deterministic Industrial Machine-to-Machine (ZT-DI-M2M) communications. We consider a scenario, in which multiple self-adaptive industrial M2M communications are deployed underlaying cellular networks to emulate a variety of vertical use cases in Industry 4.0 with the objective of guaranteeing the determinacy of latency. Specifically, in order to figure out the relationship between latency determinacy and the utilization of wireless resources meanwhile modeling the uncertainty of the stochastic environment, we firstly formulate the resource allocation problem as a stochastic game by jointly considering transmission power control, frequency spectrum allocation and the selection of base stations. Afterwards, a random graph-based sparse Long Short-Term Memory (LSTM) network is proposed to solve the optimization problem while reducing the computational complexity. Finally, numerical result demonstrates the effectiveness of the proposed scheme.