73. Skill based dynamic task allocation in Human-Robot-Cooperation with the example of welding application
ZeMA Zentrum für Mechatronik und Autom. gemeinnützige GmbH, Germany
Due to technological and organizational boundary conditions the automation of assembly processes is usually difficult to solve. It is subject to technological challenges and economical risks due to high numbers of variants of parts as well as the complexity of the assembly processes to be managed reliably. Experience has shown that full automation of production cycles is often inefficient.
For an implementation of efficient automation, it is necessary to develop flexible and adaptable production systems that can be applied to a skill-based dynamic task allocation. This ensures high efficiency of the production plants.
As part of the research projects TRSE (semi-automated robot welding for single item production) and 4by3 (Modularity, Safety, Usability, Efficiency by Human-robot-collaboration), at ZeMA seeks to develop new process technologies, planning tools, and adequate equipment in order to enable efficient and customizable automation for various production processes.
Human-robot-cooperation (HRC) is an approach of flexible automation with a skill based dynamic task allocation. Employee and robot work together without a separating protection device in an overlapping work space. The idea is to support the operator in the production process with a HRC robot system to achieve higher process efficiency and to improve quality.
One solution for a flexible skill-based automation of chosen processes is represented in Human-robot-cooperation for assembly of high quality machines for the medical industry. Instead of an unflexible fully automated approach, a semi-automated one is being developed, that can be easily controlled by the operator on the shopfloor. Therefore know-how gained from manual processes will be efficiently transferred into the assembly system.
64. Method for design of human-industrial robot collaboration workstations
1Mälardalen University, Sweden; 2Scania CV AB, Global Industrial Development, Södertälje, Sweden; 3University of Skövde, School of Engineering Science, Skövde, Sweden; 4Chalmers University of Technology, Department of Product and Production Development, Gothenburg, Sweden
Human Industrial robot collaboration (HIRC) is a rapidly growing field in research. Personal safety in a fenceless system and novel communication between human and industrial robot are two of the areas under research. The overall vision of the HIRC systems are to combine the strength, endurance and accuracy of the industrial robot with the intelligence, flexibility and tactile senses of the human to create more productive production systems with lower ergonomic loads on the operator. There already exists a number of collaborative robots on the market that is designed to stop at an unforeseen impact, thus work safe beside a human in a fenceless environment. However, these robots are typically weak and slow variants of industrial robots. Even though collaborative robots exist on the market, the possibility to virtually evaluate the entire system of robot and human collaboration, before making an investment decision, is very limited in available simulation software Even less information about methods for using simulation tools in HIRC analysis is available. In order to meet this need, a demonstrator simulation software were developed, making it possible to design and evaluate HIRC system layouts. The aim with this paper is to present a HIRC design method that put this novel demonstrator software’s possibilities in the overall production system design process. This method could be used in design decisions early in the production development process.
An industrial HIRC design case was used to identify the needs and demonstrate the use of the method. The developed HIRC design method is divided to three areas, resource allocation, simulation and evaluation of design alternatives and mathematical optimisation to achieve optimal solutions. Resource and task allocation is a challenging issue where multiple objectives have to be considered to reach the most beneficial solution. The presented method include considerations of automation constraints to limit the possible task allocations. The next area is simulation and evaluation of production designs. The simulation is based on 3D CAD data and includes design of multiple production layouts based on the geometric layout. The simulations produce quantitative outputs considering time, biomechanical load and production cost. These outputs are considered when the production system layout and the exact task allocation between human and robot are decided. Mathematical optimisation algorithms can also be used in order to find the optimal solution based on system objectives and constraints. The optimal solution to a particular production system design problem is found in the entire spectrum from a fully manual solution, a HIRC solution, to a fully automated robotic station. The presented method show how the developed demonstrator software can be used in a systematic way to enable design of productive and sustainable HIRC workstations.
283. Portable rapid visual workflow simulation tool for human robot coproduction
1Delft University of Technology, Netherlands, The; 2Zentrum für Telematik e.V., Magdalene-Schoch-Str. 5, D-97074 Würzburg, Germany
Within the European Factory-in-a-day project, the aim is to improve communication between automation integrator and factory owner, in their analysis of feasibility and appropriateness of automating a manual task. A visualisation tool with preconfigured workflows and working principles, with specific focus on efficient human-robot coproduction workflows can improve this process. This paper describes the Workflow Simulation Tool, which is part of the Human-Robot Coproduction Methodology, currently in development. The tool encompasses a portable tablet PC, which runs a visual modelling environment combined with a handheld 3D scanning solution. The tool also features pre-modelled template layouts, implementation of a checklist of persistent notes and portable visual documentation. The tool’s appropriateness was iteratively validated in collaboration with automation integrators. This evaluation showed that offering an interactive visual simulation enriches the dialogue during conceptual design and helps in revealing requirements that otherwise only appear during or after implementation.
224. Development of a dual-projected-based automated interference matrix algorithm for Industry 4.0
1Tunghai University, Taiwan; 2National Taipei University of Technology (Taipei Tech), Taiwan
In order to perform to Industry 4.0 standards, enterprises are moving forward with automatic factory and intelligent manufacturing. However, there are concerns pertaining to the use of automation for assembly sequence planning (ASP). This is a matter of great importance, as traditional product assembly incurs considerable manpower and time, generally accounting for 20–70% of the total manufacturing workload. At present, a number of systems are available that automatically generate an assembly sequence through a matrix representing the collide relationship between the components—i.e., an interference matrix. Yet, it is by no means easy to automatically define and build this kind of matrix. In this paper, we develop a dual-projected-based automated interference matrix (DPIM) algorithm that analyzes the relations between the components of a given product. In order to reduce the number of times that collide detection is performed in comparison with the method only do collision detection, the DPIM algorithm relies on static interference detection and dual-projected detection to generate a contact matrix, a direction contact matrix, and an interference matrix. By reducing the number of times collide detection is performed, DPIM can reduce the workload of assembly, thereby reducing the total manufacturing load overall and the manufacturing time likewise.