Industrial Assembly Processes Research Articles

Workflow modelling for human–robot collaborative assembly operations

With the advent of Industry 4.0 , industries are requested to provide customisation of the products, with a high number of tailored features, which a pure robotic assembly line cannot provide. This suggests using a collaborative and flexible scenario together with the need for workflow modelling architectures. These can handle the communication and interaction among the actors of the workcell to improve the collaboration, making it more fluent, reliable, and able to correct workers’ oversights. Within this scenario, the main contribution of this paper is the introduction of a general library of atomic Predicates that can be combined with a first-order logic allowing to model general industrial assembly processes in human–robot collaboration. Such logics and Predicates allow to model and supervise the workflow, remaining hidden to the operator through an interface that intervenes only when errors have to be notified. The methodology has been validated on a complex collaborative assembly use case, requiring different tools and characterized by components of several sizes and shapes. Predicates have been mainly exploited to recognize human activities, allowing modelling and supervising the workflow. • Formalization of a library of predicates for human activity recognition. • Extension of an architecture to model and supervise assembly operations workflow. • Validation of the proposed methodology in a complex use case scenario.

Real-Time Human Fault Detection in Assembly Tasks, Based on Human Action Prediction Using a Spatio-Temporal Learning Model

Human fault detection plays an important role in the industrial assembly process. In the current unstructured industrial workspace, the definition of human faults may vary over a long sequence, and this vagueness introduces multiple issues when using traditional detection methods. A method which could learn the correct action sequence from humans, as well as detect the fault actions based on prior knowledge, would be more appropriate and effective. To this end, we propose an end-to-end learning model to predict future human actions and extend it to detect human faults. We combined the auto-encoder framework and recurrent neural network (RNN) method to predict and generate intuitive future human motions. The convolutional long short-term memory (ConvLSTM) layer was applied to extract spatio-temporal features from video sequences. A score function was implemented to indicate the difference between the correct human action sequence and the fault actions. The proposed model was evaluated on a model vehicle seat assembly task. The experimental results showed that the model could effectively capture the necessary historical details to predict future human actions. The results of several fault scenarios demonstrated that the model could detect the faults in human actions based on corresponding future behaviors through prediction features.

Online performance optimization for complex robotic assembly processes

The variations and uncertainties in complex assembly processes could cause conventional industrial robots failure to perform assemblies. Therefore, intelligent robots must be developed which can adapt to these variations and uncertainties; moreover, the assembly process parameters must be optimized to satisfy the performance requirements such as First Time Throughput (FTT) rate and cycle time. However, it is challenging to optimize the performance of complex robotic assembly systems. The existing solutions are not efficient and human knowledge is needed to perform experiments and explore optimal parameters. These requirements greatly limit the applications of robotic assembly in manufacturing automation. In this paper, a robotic assembly process optimization method is developed for complex assembly processes. We propose a modeling method based on Gaussian Process Regression (GPR) to construct a model to build the relationship between the process parameters (input) and system performance (output). The GPR surrogated Bayesian Optimization Algorithm (GPRBOA) is improved to iteratively optimize the robotic assembly performance. Two industrial assembly processes (a tight-tolerance peg-in-hole assembly process and a torque converter assembly process) are used to verify the proposed method. The experiments were performed many times and the results demonstrate that the proposed GPRBOA method can optimize the complex assembly process parameters effectively and efficiently. The proposed complex assembly process optimization method opens a door for online manufacturing process optimization and will greatly reduce the manufacturing cost.

Decision-Making Framework for Implementing Safer Human-Robot Collaboration Workstations: System Dynamics Modeling

Human-Robot Collaboration (HRC) systems are often implemented seeking for reducing risk of Work-related Musculoskeletal Disorders (WMSD) development and increasing productivity. The challenge is to successfully implement an industrial HRC to manage those factors, considering that non-linear behaviors of complex systems can produce counterintuitive effects. Therefore, the aim of this study was to design a decision-making framework considering the key ergonomic methods and using a computational model for simulations. It considered the main systemic influences when implementing a collaborative robot (cobot) into a production system and simulated scenarios of productivity and WMSD risk. In order to verify whether the computational model for simulating scenarios would be useful in the framework, a case study in a manual assembly workstation was conducted. The results show that both cycle time and WMSD risk depend on the Level of Collaboration (LoC). The proposed framework helps deciding which cobot to implement in a context of industrial assembly process. System dynamics were used to understand the actual behavior of all factors and to predict scenarios. Finally, the framework presented a clear roadmap for the future development of an industrial HRC system, drastically reducing risk management in decision-making.

Montaže u čovek-robot kolaborativnoj radnoj ćeliji pomoću učenja pojačanjem - Reinforcement Learning

Human-Robot Collaborative (HRC) workcells could enhance the inclusive employment of human workers regardless their force or skills. Collaborative robots not only substitute humans in dangerous and heavy tasks, but also make the related processes within the reach of all workers, overcoming lack of skills and physical limitations. To enable the full exploitation of collaborative robots traditional robot programming must be overcome. Reduction of robot programming time and worker cognitive effort during the job become compelling requirements to be satisfied. Reinforcement learning (RL) plays a core role to allow robot to adapt to a changing and unstructured environment and to human undependable execution of repetitive tasks. The paper focuses on the utilization of RL to allow a robust industrial assembly process in a HRC workcell. The result of the study is a method for the online generation of robot assembly task sequence that adapts to the unpredictable and inconstant behavior of the human co-workers. The method is presented with the help of a benchmark case study.

Engineering Human–Machine Teams for Trusted Collaboration

The way humans and artificially intelligent machines interact is undergoing a dramatic change. This change becomes particularly apparent in domains where humans and machines collaboratively work on joint tasks or objects in teams, such as in industrial assembly or disassembly processes. While there is intensive research work on human–machine collaboration in different research disciplines, systematic and interdisciplinary approaches towards engineering systems that consist of or comprise human–machine teams are still rare. In this paper, we review and analyze the state of the art, and derive and discuss core requirements and concepts by means of an illustrating scenario. In terms of methods, we focus on how reciprocal trust between humans and intelligent machines is defined, built, measured, and maintained from a systems engineering and planning perspective in literature. Based on our analysis, we propose and outline three important areas of future research on engineering and operating human–machine teams for trusted collaboration. For each area, we describe exemplary research opportunities.

A Capability-Aware Role Allocation Approach to Industrial Assembly Tasks

The deployment of industrial robotic cells based on lean manufacturing principles enables the development of fast-reconfigurable assembly lines in which human and robotic agents collaborate to achieve a shared task. To ensure the effective coordination of the shared effort, each task must be decomposed into a sequence of atomic actions that can be assigned either to a single agent or to the combination of more agents, according to a defined metric. While task allocation is a general problem and has been discussed intensively in other fields, less effort has been devoted in industrial scenarios, involving mixed human–robot teams and in particular, to the factors that should be considered in allocating tasks among a heterogeneous set of agents in collaborative manufacturing scenarios. In this letter, we investigate the agent characteristics that should be considered in the task allocation problem of fast-reconfigurable systems in industrial assembly processes. First, we introduce a set of indices, namely task complexity, agent dexterity, and agent effort, to evaluate agent performance with respect to a task. Second, we propose an offline allocation algorithm that combines the performance indices to assign optimally the task to the team agents. Finally, we validate the framework in a proof-of-concept collaborative assembly of a metallic structure. The results show that the workload is shared through the agents according to their particular physical capabilities and skill levels. A subjective analysis of the proposed collaborative framework on 12 healthy participants also validated the intuitiveness-of-use and improved performance.

Vision-Based Tip Position Control of a Single-Link Robot Manipulator

Nowadays, robot manipulators are an important part of the automation and control sector, so they indicate the significance of the associated control methodologies. These finds broad applications in various areas, especially in medical, aerospace and industrial assembly processes. Rigid robot manipulators form an important class of manipulators where they interact with the environment physically. Robot manipulator is a non-collocated system that leads to inaccurate system performance. To solve the aforesaid issue, a vision sensor can be employed for direct measurement of tip position instead of the traditional mechanical sensors. The major objective of the work presented here is development of a vision-based method for controlling the tip position of a planar 2-DOF single-link rigid robot manipulator (2DSLRM) from one point to another. In this paper, to achieve the desired tip position, feedback is taken from an encoder and visual sensor from the tip of Single-Link Robot Manipulator (SLRM). The acquired results show that the developed control method improves tip position control performance of 2DSLRM as compared to the traditional mechanical sensors. Simulation and experiment have been performed to study the performance of the proposed tip position control of 2DSLRM.

Implementation of virtual reality systems for simulation of human-robot collaboration

A collaboration between human and robot can implicate many advantages for complex industrial assembly processes, especially as increased flexibility and adaptability become a key feature of production systems. The use of virtual reality (VR) systems has the potential to simulate cooperative processes in advance and to include workers and their individual behavior into the simulation. The use of VR simulations makes it possible to secure processes and reduce physical and mental barriers between human and robot. This paper presents a methodical approach for the implementation of systems for the virtual testing of collaborative assembly processes. Following the aim of replicating the assembly process with the highest possible immersion, a specific VR system is derived from an analysis of the assembly process. Core features of the system are the physical simulation of the assembly process, the integration of the robot control and a haptic feedback for the operator.

Metal muscles and nerves—a self-sensing SMA-actuated hand concept

Bio-inspired hand-like grippers actuated by Shape Memory Alloy (SMA) wires represent an emerging new technology with potential applications in many different fields, ranging from industrial assembly processes to biomedical systems. The inherently high energy density makes SMAs a natural choice for compact, lightweight, and silent actuator systems capable of producing a high amount of work, such as hand prostheses or robotic systems in industrial human/machine environments. In this work, a concept for a compact and versatile gripping system is developed, in which SMA wires are implemented as antagonistic muscles actuating an artificial hand with three fingers. In order to combine high gripping force with sufficient actuation speed, the muscle implementation pursues a multi-wire concept with several 0.1 mm diameter NiTi wires connected in parallel, in order to increase the surface-to-volume ratio for accelerated cooling. The paper starts with an illustration of the design concept of an individual 3-phalanx-finger, along with kinematic considerations for optimal placement of SMA wires. Three identical fingers are subsequently fabricated via 3D printing and assembled into a hand-like gripper. The maximum displacement of each finger phalanx is measured, and an average phalanxes dynamic responsiveness is evaluated. SMA self-sensing is documented by experiments relating the wires change in resistance to the finger motion. Several finger force measurements are also performed. The versatility of the gripper is finally documented by displaying a variety of achievable grasping configurations.

A Hierarchical Human-Robot Interaction-Planning Framework for Task Allocation in Collaborative Industrial Assembly Processes

A Hierarchical Human-Robot Interaction-Planning Framework for Task Allocation in Collaborative Industrial Assembly Processes

Status of Light Weight Cassette Design of SOFC

Lightweight SOFC stacks are currently being developed for stationary applications such as residential CHP units, for automotive applications such as APU and for portable devices. Within the EU funded project the so-called Jülich CS-design has been improved for lightweight SOFC stacks using glass-ceramic sealings that decouple the thermal stresses within the stack and at the same time allows optimal sealing and contacting. The design is highly suitable for industrial low-cost manufacturing and automated assembly. To determine the effect on the thermomechanical behaviour of the fuel cell stack, modelling and coupled thermomechanical analysis have been performed. Based on the analysis results in combination with the manufacturing experience, test results, and post-test analysis from previous stack design, substantial changes have been made to the CS-design in order to improve mechanical robustness of the stack. The manufacturing of single parts, particularly due to the improved design of sheet metal interconnects, as well as the assembling processes are suitable for low-cost mass manufacturing. The novel decal concept of glass-ceramic sealant screen printed on foil in order to produce green tapes is used for joining the stack layers offering an enormous potential for cost savings in industrial assembly process. The new design CSV(D2.0 in the MMLCR=SOFC project) furthermore has extended the process window of the laser welding joining process. First CSV–design stack tests showed a comparable electrochemical performance to the previous CSIV design. The good adhesion of the glass sealant applied on steel sheet by screen printed tape and good contacting behaviour between cells and interconnects were furthermore confirmed by post-test analysis.

Status of Light Weight Cassette Design of SOFC

Lightweight SOFC stacks are currently being developed especially for automotive applications such as APU and for portable devices. Within the EU funded project MMLCR=SOFC the Jülich lightweight so-called CS-design was improved concerning better suitability for glass sealing, reduced manufacturing effort and increased power. Based on modelling in combination with manufacturing experience, test results, and post-test analysis substantial changes of the previous CS-design were made. The manufacturing of single parts, particularly due to the improved design of sheet metal interconnects, as well as the assembling processes are suitable for low-cost mass manufacturing. The novel decal concept of glass-ceramic sealant screen printed on foil in order to produce green tapes is used for joining the stack layers offering an enormous potential for cost savings in industrial assembly process. First stack tests with the new CSV–design showed a comparable electrochemical performance to the previous CSIV design having at the same time a better thermo-mechanical behavior.

Welding Simulation of Non-Nominal Structures With Clamps

In any industrial assembly process, there are a number of sources of variation. Variation in the manufacturing process leads to component variation, which, together with fixture variation and variation stemming from the joining process, propagates to the final product. In order to analyze and diminish the effect of variation, it is important to identify and be able to simulate the phenomena contributing to final variation. In this paper, the focus is variation in welding distortion arising from non-nominal components that are joined. In the welding process, it has been shown that variation in components and in fixtures influences the size and distribution of weld-induced distortion. Hence, in order to accurately simulate geometric variation of an assembly joined by weld joints, variation simulation and welding simulation need to be performed in combination. Previous research that focused on the combination of variation simulation and welding simulation has not considered components that are clamped. Instead the components were treated as rigid bodies at non-nominal positions prior to welding. In many industrial applications, clamps are used when assemblies are welded, and it is therefore important to quantify the influence that clamping has on welding of non-nominal components. In this paper, we simulate the combination of variation in components and fixtures with welding, considering that the components are clamped prior to welding. Although clamps will force the components closer to their nominal positions along the weld joint, they also introduce a stress field in the structure, which together with the welding process may cause additional distortion. Two case studies are performed and analyzed: a T-weld joint and a butt-weld joint. The results show that welding distortion depends on fixture error even in the presence of clamps.

Human-robot Cooperation in Future Production Systems: Analysis of Requirements for Designing an Ergonomic Work System

Both humans and robots have crucial advantages regarding industrial assembly processes. While robots ace at repetitive and monotonous assembly steps, humans are able to adapt flexibly to new situations and upcoming problems. Combining these advantages by means of direct human-robot cooperation seems to be interesting for producing companies but has not been realized yet. While the corresponding technologies are already available, appropriate safety standards to ensure occupational safety are missing and represent one of the main barriers for introducing direct human-robot cooperation. This paper describes the requirements for a workplace, where humans and robots jointly perform an assembly process without separation between their workspaces. The requirements are identified in line with an ergonomic workplace for different aged working persons, whereas the robot assists the human with the assembly process. The analysis considers technical, human-related, and normative requirements. Afterwards, an early implementation of the concept is presented based on a cognitively automated assembly system using a lightweight robot.

Human-Robot Interaction Based on Use of Capacitive Sensors

In industrial assembly processes nowadays there is a trend of incorporating robotics and automation in more fields of human activity. The objective is to alleviate the work process for human operators and improve operations with the ultimate goal of increasing precision, shortening the cycle time and reducing risks. However current limitations of technical systems still dictate that human operator takes part in the process and also provides necessary support to the system. Development of advanced interaction models could simplify the role of human operators mainly in complex tasks. Robotic assistants, in form of an industrial robot, completely replace human operators in processes they are fit to perform or assist and hand items to them when they are not. When working together the robot’ priority is not to harm the human. Also it is important the human can convey desired actions of the robot in an intuitive manner. The goal is to apply human robot interaction models based on tactile stimulus. A simple capacitive sensor field has been developed for aiding the human-robot interaction in industrial applications. The sensor is connected to Arduino controller and mounted on the robot. When the human operator is in contact with the robot the change of capacitance is detected. The robot responses appropriately based on the input.

Cross-disciplinary method for predicting and reducing human error probabilities in manual assembly operations

Currently, methods for the evaluation of human reliability are mainly applied in safety critical industries (aerospace engineering, nuclear engineering). The Expert System for Task Taxonomy (ESAT) was developed to evaluate the error rate of control and surveillance activities in safety critical areas. In contrast, currently industrial assembly processes are mainly developed concerning an optimal use of the time factor. In this article, the suitability of ESAT for the evaluation of human reliability in manual assembly operations will be examined. By the analysis of an exemplarily considered assembly process it is shown whether and how its work content can be depicted by the ESAT method. In this context, it is also analysed how a modification of the marginal conditions – e.g. the level of specification – affects the calculation results. In the first analysis six manual assembly operations from the automotive industry were analysed and appropriate ESAT-data for its sequences of action were calculated. As a result, the new method allows quantifying potential human error rates before the start of production and comparing planning alternatives under time and cost aspects early. Consequently, the evaluation of human mistakes during the planning process allows developing not only time-optimised but also quality-optimised manual assembly operations.

AI tools for use in assembly automation and some examples of recent applications

PurposeThis paper aims to review seven artificial intelligence tools that are useful in assembly automation: knowledge‐based systems, fuzzy logic, automatic knowledge acquisition, neural networks, genetic algorithms, case‐based reasoning and ambient‐intelligence.Design/methodology/approachEach artificial intelligence tool is outlined, together with some examples of their use in assembly automation.FindingsArtificial intelligence has produced a number of useful and powerful tools. This paper reviews some of those tools. Applications of these tools in assembly automation have become more widespread due to the power and affordability of present‐day computers.Research limitations/implicationsMany new assembly automation applications may emerge and greater use may be made of hybrid tools that combine the strengths of two or more of the tools reviewed in the paper. The tools and methods reviewed in this paper have minimal computation complexity and can be implemented on small assembly lines, single robots or systems with low‐capability microcontrollers.Practical implicationsIt may take another decade for engineers to recognize the benefits given the current lack of familiarity and the technical barriers associated with using these tools and it may take a long time for direct digital manufacturing to be considered commonplace… but it is expanding. The appropriate deployment of the new AI tools will contribute to the creation of more competitive assembly automation systems.Social implicationsOther technological developments in AI that will impact on assembly automation include data mining, multi‐agent systems and distributed self‐organising systems.Originality/valueThe novel approaches proposed use ambient intelligence and the mixing of different AI tools in an effort to use the best of each technology. The concepts are generically applicable across all industrial assembly processes and this research is intended to prove that the concepts work in manufacturing.

Experience-based optimization of universal manipulation strategies for industrial assembly tasks

The trend towards smaller lot sizes and shorter product life cycles requires automation solutions with higher flexibility. Today’s robotic systems often are uneconomical for frequently changing boundary conditions and varying tasks due to high engineering costs needed for a well-defined supply of parts and pallets. At the same time, even small inaccuracies due to shape deviations in parts or pallets often cause high downtime. This work contributes to the robustness of industrial assembly processes with high inaccuracy concurrent to narrow tolerances. Therefore, contact-based manipulation strategies are defined, which are model-free and object-independent and solve common industrial tasks as palletizing, packaging and machine feeding. While the strategies are robust to inaccuracy up to 5 mm/5° due to localization uncertainty or object displacement, they handle usual industrial assembly tolerance of far below 1mm. The necessary flexibility and reusability for new tasks is guaranteed by hierarchical decomposition into atomic sub-strategies. In order to accelerate the execution, the manipulation strategies are customized to each specific task by unsupervised experience-based learning. The flexibility of the manipulation strategies and the progress in cycle time during the execution are shown on common industrial tasks with varying objects, tolerances and inaccuracies.

Potential health and environmental issues of mercury-contaminated amalgamators

Dental amalgamators may become contaminated internally with metallic mercury. This contamination may result from mercury leakage from capsules during trituration or from the long-term accrual from microscopic exterior contaminants that result from the industrial assembly process. The potential health risk to dental personnel from this contamination is unknown. The authors assessed used amalgamators from the federal service inventory for the amounts of mercury vapor levels, as well as the visual presence of mercury contamination. They evaluated these amalgamators for potential mercury vapor health risk, using established National Institute for Occupational Safety and Health methods and American Conference of Governmental Industrial Hygienists standards. Ten of the 11 amalgamators assessed had measurable mercury vapor levels. Four amalgamators were found to have internal static mercury vapor levels above Occupational Safety and Health Administration ceiling limit thresholds. During a simulated worst-case clinical use protocol, the authors found that no amalgamators produced mercury vapor in the breathing space of dental personnel that exceeded established time-weighted federal mercury vapor limits. Amalgamators may be contaminated internally with metallic mercury. Although the authors detected mercury vapor from these units during aggressive, simulated clinical use, dilution factors combined with room air exchange were found to keep health risks below established federal safety thresholds. Dental personnel should be aware that amalgamators may be contaminated with mercury and produce minute amounts of mercury vapor. These contaminated amalgamators may require disposal as environmentally hazardous waste.