Safe Collaborative Assembly on a Continuously Moving Line with Large
Industrial Robots
Varun Gopinath
⁎,1,a, Kerstin Johansen
a, Micael Derelöv
a, Åke Gustafsson
b, Stefan Axelsson
baLinköping University, Department of Management and Engineering, Sweden bVolvo Car Corporation, Göteborg, Sweden
A R T I C L E I N F O Keywords: Collaborative Operations Risk Assessment Industrial Robots Hazards Assembly operations Continuously moving line
A B S T R A C T
Robot safety standards defines Collaborative Operation as a state in which purposely designed robots work in direct cooperation with a human within a defined workspace. That is, an operator and an industrial robot complete assembly tasks at the collaborative workspace. A prerequisite to ensuring safety during all phases of operation is an understanding of the nature of hazards pertinent to collaborative systems. An automotive as-sembly station, where plastic panels are assembled on a continuously moving line, formed the basis for research operations meant to understand safety issues when a large industrial robot aids an operator in assembly tasks. This led to the development of a laboratory demonstrator whose design and functioning will be presented in this article. Additionally, the hazards identified during risk assessment along with measures to mitigate the asso-ciated risks will be presented in order to highlight the nature of hazards pertinent to collaborative systems.
1. Introductions
Recent years have witnessed advances in several technologies such as in the design of robots, control systems as well as sensors that have made it possible for industrial robots to be used safely withinfinal as-sembly lines[1,2]. This is referred to as Collaborative Operation[3,4], whose functional principle is to combine the characteristics of robots with the superior cognitive and decision-making skills of an operator. That is, in Collaborative Operation[5], industrial robots and operators work together at the collaborative workspace performing tasks referred to as collaborative tasks. The expected outcome of collaborative as-sembly is an improvement in the efficiency in asas-sembly operations which are currently carried out manually.
Automotive plants are characterized by: 1. high-volume production that motivates continously moving lines for part transfer between sta-tions; 2. manual-labor which implies design of assembly plants that cater to requirements of operators. Therefore, the challenge of safely introducing robots in a populated environment is currently a research interest[2,6]. There are several perceived benefits to Collaborative Operation such as: 1. Ability to adapt to marketfluctuations and trends [7,8], 2. Decrease cycle time[9]and 3. Improve working environment [10,11].
The close proximity of humans and robots in Collaborative Operation implies higher probability for occurrence of hazardous
situations[4,12–15].Therefore, a proper understanding of the nature of hazards are of particular importance in order to benefit from the per-ceived benefits of Collaborative Operation[16,17]. Currently, safety issues with regard to robots specifically designed for collaborative work, referred to as collaborative robots have been of significant in-terest to both engineers and researchers[18,19]. However, large or traditional industrial robots (e.g., KUKA KR-210 [19]) have higher performance in tems of payload and reach than collaborative robots. Therefore, insights into safety issues related to large robots could ad-vance Collaborative Operation in manufacturing plants.
Ericson[20]notes that identification of hazards is a critical and non-trivial process, whereas the solutions to mitigate the risks are re-latively straightforward. Additionally, Etherton et al. [21]notes that designers lack a database of known hazards during innovation and design stages. Also, robot safety standards (ISO 10218-1/2[3,22]) have tabulated a list of significant hazards whose purpose is to inform risk assessors of common hazards associated with robotic systems. However, there is a lack in knowledge of the nature of hazards in academic lit-erature and therefore this article aims to contribute to this under-standing by investigating safety related issues pertinent to Collabora-tive Operation with large industrial robots.
This article expands upon research presented by Gopinath et al.[23] and presents a physical demonstrator developed in a laboratory en-vironment to showcase: 1. Design and functioning of a safe
https://doi.org/10.1016/j.rcim.2020.102048
Received 9 October 2019; Received in revised form 31 July 2020; Accepted 31 July 2020
⁎Corresponding author.
E-mail address:varun.gopinath@liu.se(V. Gopinath).
1www.liu.se
0736-5845/ © 2020 Published by Elsevier Ltd.
application of current safety techniques and practices to ensure safe operations. Additionally, the demonstrator will be used to highlight the nature of hazards pertinent to collaborative operations with large in-dustrial robots.
This article is structured as follows: Section: 2 presents relevant theoretical background followed by a brief presentation of the research method insection: 3. A case study is presented insection: 4which forms the basis for the laboratory demonstrator described insection: 5. The nature of the collaborative workstation is discussed insection: 6with final conclusions insection: 7.
2. Theoretical Background
This section presents a theoretical overview of assembly work-stations followed by a presentation on industrial safety, collaborations operations and machine interfaces.
2.1. Assembly Workstation
An assembly workstation functions to assemble components that constitute a product such as a vehicle [24]. They are discrete work-spaces and according to Hales (ch. 40[25]), physical, procedural and personnel are three aspects that must be addressed when planning an assembly cell. They consist of infrastructure such as workspaces, part handling systems, machinery, and devices to display information. Pre-defined tasks are carried out by workers and are allocated to specific workstations that require operating procedures for efficiency, quality, maintenance etc.
The assembly of an automotive begins with the painted chassis and continues along the line where parts are added sequentially[26]. The movement of the assembly along the line can be carried out with in-termittent or continuous motion. Two reasons are attributed to the higher cost of a continuously moving line: 1. It requires a robust and automated part handling system that moves the chassis along the line. 2. It also demands handling of greater quantity of parts that will be assembled due to the lower cycle time associated with such systems [25]. Assembly operations are traditionally manual-labour intensive which implies that the design of assembly workstations prioritize re-quirements of workers. These include: 1. shorter distance between workstations which facilitate communication and interaction. 2. easy entry and exit from workstations. 3. reduced requirements on handling, cycle times, inventory as well as physical space compared to automated solutions (ch 40[25]).
2.2. Industrial Safety
Theoretical aspects related to safety followed by current safety practices in industrial operations will be presented.
2.2.1. Hazard Theory
An accident is an event which occurs when a hazard, or more spe-cifically a hazardous element, is activated by an initiating mechanism
defined as a potential source of harm[27]and is composed of three basic components: 1. Hazardous Element (HE) is a resource that has the potential to create an accident (e.g fast moving robot). 2. Initiating Mechanism (IM) is a trigger or initiator event causing the hazard to occur (e.g. obstruction in robot path). 3. Target/Threat (T/T) are the persons or the equipment directly affected when an accident occurs. As noted by Levesson[28], a hazard is designed into a system and the occurrence of an accident depends on two factors: 1. unique set of hazard components and 2. accident risk represented by these compo-nents, whereRisk=Probability×Severity.
2.2.2. Risk Assessment and Risk Reduction
The machine safety standard [27] suggest carrying out Risk As-sessment (RA) followed by implementing Risk Reduction measures to ensure the safety of the operator along with manufacturing processes (fig: 1). RA is an iterative process that concludes when all probable hazards have been identified along with solutions to mitigate the effects of these hazards have been implemented.
Risk Assessment consists of risk analysis followed by risk evaluation (fig: 1). The risk analysis process is as follows: 1. Determine the limits of the machinery: The aim is to specify the overall operating environment of the robotic system including different phases of machine operation along with the tasks to be executed. The tasks can be formulated in terms of robot, operator and collaborative tasks[17]. 2. Hazard-Iden-tification: This is the critical step of identifying the hazards associated with each task. 3. Risk Estimation: The risks associated with each hazard are estimated, which is a function of the severity of harm and the probability of occurrence of that harm. When the risks are estimated, the critical step of evaluating the risk (risk evaluation) which aims to determine if the effect of the hazard warrant a risk reduction measure. To effectively manage risks, the designer has the choice of im-plementing safe solutions through three steps: 1. Inherently safe design measures: refers to design changes that eliminate the source of hazards; 2. Safeguarding and/or complementary protective measures: refers to in-stalling additional solutions such as physical guards and safety sensors; 3. Information for use: includes training, documentation and other means that attempts to eliminate residual risks.
Risk assessment is carried out by a team participating in workshops with helpful aids such as sketches, pictures or visits to the installation. As noted by Marvel and Norcross[14], the discussions are held in plain language in order to describe hazardous situations and the interpreta-tion of relevant standards. Ericson[20]) describes hazard identification as a cognitive process where the team have to document foreseeable hazards for the system under investigation. Hazardous situations are identified for a task performed by the system or the operator. The team can refer to a list of common hazards documented in ISO 10218-1/2 along with additional hazards can be postulated based on the working environment.
2.2.3. Safeguarding and System Architecture
Physical fences are used to safeguard operators from accidentally entering a hazard zone. The entry to a hazard zone is protected with
interlocking devices that can trigger a safety function if an entry is detected. Sensitive Protective Equipment (SPE) are sensors to monitor hazardous areas in order to trigger a safety function upon detection of an intrusion at the safeguarded zone. The selection, positioning, con-figuration and commissioning of SPE’s are standardized in order to achieve the required performance level. IEC 62046 [29]defines the performance requirements for these equipment and as stated by Marvel and Norcross[14], when triggered, these sensors use electrical safety signals to trigger safety function of the system. Examples include Electro-sensitive protective equipment (e.g., Light curtains & laser scanners) along with pressure-sensitive protective equipment.
Industrial robots are integrated with machinery that forms a robotic system/workstation. A simplified system architecture of a typical ro-botic system is shown infigure: 2. A safety controller consists of a logic processing unit, coupled to one or more Input/Output modules, safety relays, communication modules etc. (Ch.54[2]). A safe logic processing unit (similar to a PLC) is designed with features such as redundant cpu, ability to communicate with safety sensors, etc. They can function as the master controller, where the subsystems are manufacturing ma-chines, robots, sensors, and interface devices. The master controller is used to coordinate action between the devices during the entire cycle. They are used for various safety related functions such as signalling safety relay that removes power from machines, co-coordinating robot motion and communicating with external systems in the plant. 2.2.4. Collaborative Operations
Robot safety standards (ISO/EN: 10218-1/2 [3,22] and ISO/ TS 15066[5]) presents four types of collaborative operations and notes that a collaborative workstation shall comply with one or more of these types. The four types of collaborative operations are: 1. Safety-rated monitored stop; 2. Hand Guiding; 3. Speed and separation monitoring and 4. Power and force limiting Collaborative operation is defined as a state in which a purposely-designed robot system and an operator work within a collaborative workspace [5]. The implementation of Colla-borative operation can be understood as a mode of operation (see sec: 2.3), where a robot and an operator complete tasks at the colla-borative workspace.
2.3. Interfaces for Human-Robot Collaboration
Collaboration is an act of doing something with someone to achieve a goal, whereas interaction refers to the act of communication between two participants that intended to collaborate. Sarter[30]notes that in complex human-automation collaborative systems, feedback between the human and the automation device should be planned and designed to avoid hazardous situations. Interfaces can support operators under several situations such as: 1. improve situational awareness by commu-nicating the state of the system and 2. support awareness of the current operating mode (mode-awareness) of the system.
Production facilities employ tools such as warning lamps, buttons, enabling devices,floor marking etc., to provides situational awareness and control to the operator[31,32]. Warning lamps andfloor markings are visual tools to support awareness of the system and its boundary whereas enabling devices are tools to safely initiate an automation function.
Computer-based technologies permit designers of automation system to develop systems that can change its behaviour and is referred to as mode of operation. According to Sarter and Woods [30], in complex systems, operators needs to: 1. keep track of the current mode; 2. know when and how to change the mode; 3. understand the function of each mode. They note that improperly designed systems can increase the cognitive demands on the operator. Therefore, maintaining mode-awareness can be particularly challenging when a system changes its behavior based on environmental input or for protection purposes. Si-tuational and mode awareness of the system can be considered by the risk assessment team (sec: 2.2.2) as potential hazard (sec: 2.2.1) that could lead to an accident or unplanned stop[33].
3. Methodology
The research methodology[34,35]is illustrated infigure: 3where activities directed towards understanding the problem (sec: 3.1) over-laps with those directed towards the development of demonstrators (sec: 3.2).
3.1. Problem Study and Data Collection
To understand the safety requirements of collaborative operations, a case based approach[36]was employed and was focused on an auto-motive assembly station as detailed insection 4. The following methods to understand requirements were employed: 1. Detailed study of the station by observation and participating in the assembly process. 2. Regular meetings with engineers and line-workers to gain insight into their work environment [37]. 3. Literature sourced from academia, books as well as documentation from various industrial equipment manufacturers were reviewed. 4. Participation in trade-shows and conferences permitted the researchers to document the state of the art in robotic technologies and trends. 5. Training in risk assessment from authorized organizations supported interpretation of the standards and industrial practices.
3.2. Demonstrator Development
Demonstrator-Based Research (DBR) is a multi-methodological ap-proach to research developed to address challenges faced by a manu-facturing organisation[35,38,39]. Björnsson[39]notes that a demon-strator can serve two purposes: 1. A platform for experimentation to build new theories and gain insights into issues related to research Fig. 2. A simplified architecture of a robotic workstation where a safety controller coordinates safety-related functions.
objectives and 2. A communication tool between the researcher and the industry. Jonsson[38]and Bjornsson[39]also note that demonstrator can be developed to many forms such as prototype, virtual or physical demonstrators.
The development of demonstrators can be a path to knowledge discovery and can provide insight into issues that motivates research in the first place. As motivated by Blessing and Chakrabarti[40], con-ceptualization of solutions– in the form of demonstrators – can give rise to questions about the nature of the problem. This encourages re-searchers to plan parallel activities that attempts to address it. That is, development of demonstrators (fig: 3) is iterative where each phase increases knowledge and confidence in the solution.
4. Case study– Under-body assembly on a continuously moving line
To improve aerodynamic performance and fuel efficiency, light, and flexible plastic panels (less than 1kg) are installed under the body of a vehicle. There are many panel variants and the tasks are performed by one operator manually. The plan-view of the assembly cell where the assembly is continuously moving (100mm/sec) is illustrated infigure: 4. The task sequence can be divided into two: 1. Selecting andfixing the panel under the car. 2. Securing the panel with bolts using the nut-runner.
The cycle time of 1min (approx.) and the sequence of tasks can be elaborated as follows: 1. Identify the vehicle model by reading the in-formation label; 2. Select and installs clips on the correct panel; 3. Fix the panel under the car; 4. Pick up the pneumatic nut-runner and 5. Fasten 8-10 screws to secure the panel. The automotive industry ex-pects an increase in the number of panels per car in future vehicles. In order to investigate requirements on safety in collaborative operations, the tasks can be conceptualized in either one of two ways: 1. The robot picks the panel and places it under the car while the operator secures the panel with bolts (sec: 5). 2. The operator picks the panel and places it under the car moving continuously while the robot fastens bolts to secure the panel[41].
5. Collaborative Assembly on a Continuously Moving Line The tasks (sec: 5.1) and the layout (sec: 5.2) of a demonstrator that exemplifies a safe workstation where a large robot and an operator collaborate to assemble under-body panels on a continuously moving line (fig: 5) is presented.
5.1. Operator, Robot and Collaborative Tasks
Figure: 6presents the sequence of tasks carried out by the operator to assemble the plastic panels on a simulated car body. Operators and robots are two resources allocated to the collaborative workstation. The tasks are delegated between these two resources and are defined in terms of robot (RT), operator (OT) and collaborative (CT) tasks. Tasks carried out by operator and robot are OT and RT respectively, whereas tasks performed in the collaborative workspace (CW) are termed col-laborative tasks. The task of picking the panel and placing it under the moving line is delegated to the robot and the operator is tasked with fastening the panel with screws. The task-sequence is as follows:
1. OT/RT: Standing at the operator workspace (fig: 5), the operator monitors the workstation while preparing for the upcoming as-sembly task. The operator monitors the system directly or the warning lamps. In automatic mode, the robot moves at high speed to pick up the panel (fig: 6a).
2. RT: The robot picks the panel using suction cups (fig: 6b), proceeds to move to the collaborative workspace (fig: 5). The robot syn-chronises with the moving line and initiates collaborative mode (fig: 6c).
3. CT: When collaborative mode is initiated, the lamp turns green and laser scanner (fig: 5) mutes monitoring the CW. The operator enters the CW and begin fastening screws to secure the panel (fig: 6d). 4. OT: When the screws are fastened, the operator goes out of the CW
(fig: 6e) and engages the mode-change button (fig: 6f). The mode-change button (fig: 5) blinks when the line is at 80% of its travel (not shown), in order to direct operator’s attention.
5. RT: Then the laser scanner initiates monitoring of CW, turns the lamp from green to red and moves back to the robot workspace and begins the next cycle (fig: 6f).
5.2. Layout of the Collaborative Workstation
In order to simulate a continuously moving line, a structure (fig: 5) was designed to support a linear actuator (fig: 5). A belt driven ELGA-TB-RF-70-3000-0H linear actuator with a CMMO-ST-C5-1-DIOP [43] motor controller was integrated with the robotic system. A bracket on the linear actuator approximated the mounting holes of the car body. The linear actuator moves at a constant speed of 100mm/sec for Fig. 3. Illustration of research methodology developed to understand safety-related issues in collaborative operations with large industrial robots. Adapted from Gopinath et al.[34].
Fig. 4. Description of the tasks performed by an operator from reading the information label to fastening the panel with screws for one car.
3000mm.
Workspaces and safeguarding strategies Three distinct workspaces– Robot, Operator & Collaborative (fig: 5)– together form the boundary of the workstation and safeguarded using physical fences and safety sensors. Fences physically separate the robot workspace (RW) and the collaborative workspace (CW) and are installed to eliminate the risk of an operator entering a hazard zone.
A laser scanner (fig: 5) monitors the robot and collaborative work-spaces. A light curtain (fig: 5) mounted at one extremity of the station was installed to take into account an unforeseeable incident that results in the operator unable to trigger the change button. The mode-change button initiates the next assembly cycle and is part of the in-terface cluster shown infigure: 5.
Figure: 7shows the robot following the moving line. The illustrated volume is the limits within which the end-effector can be programmed to move within the collaborative workspace. That is, in collaborative mode, a safe volume is triggered and until automatic mode is initated (with mode-change button) by the operator, the robot moves within this volume at the programmed speed.
End-effector Suctions cups are used to grip the plastic panels be-cause: 1. They can grip a part from one side and 2. They can adapt to minor changes in part location (in the storage rack) and geometry. The work process states (sec: 5.1) that the end-effector moves through the
Fig. 5. Picture of the laboratory demonstrator developed to showcase safe collaborative assembly on a moving line.
Fig. 6. Illustration of the sequence of tasks to collaboratively assemble a plastic panel under the body of a (simulated) car[35,42].
Fig. 7. Illustration of a spatial volume (marked as Monitored Space) within which the robot is permitted to move during collaborative mode.
the CW is low, it is necessary to avoid proximity to sharp edges. Therefore, a soft cover was designed that hangs on the end-effector and can deflect 50mm (approx.) in all directions.
Hazards and risk reduction Risk assessors are tasked with identifying hazards that can lead to an accident. The hazards are analyzed and evaluated to warrant a change in design or inclusion of safe guarding measures.Table: 1presents the hazards identified during Risk Assess-ment. The hazards are presented in terms of the hazard components (seesec: 2.2.1) along with measures that suitably mitigate risks asso-ciated with these hazards.
5.3. Control Architecture
Figure: 8 illustrates the safety architecture of the demonstrator presented infigure: 5. The master safety controller coordinates signal flow between various devices shown in figure: 8. The master safety controller consists of 1. Safety CPU, 2. Power supply, 3. Input/Output (I/O), 4. Safety relays and 5. networking modules. In the demonstrator, the Input/Output modules were programmed to set the correct signal to the safety lamps and receive inputs from the human-interface devices. The human-interface devices are emergency buttons, mode-change button, reset/clear safety signals.
The master controller receives signals from the safety sensors and can be programmed to mute or set the monitoring fields.Figure: 8 shows the fields monitored by the laser scanner where the left side corresponds to the CW while the right side corresponds to the RW. The gap in the monitoringfield is due to the legs of the physical fences that forms a shadow and is a potential blind spot for the laser scanner. It is
of operation. For example, during collaborative mode, the sensor monitors the RW and during other phases of the assembly cycle, the system monitors both CW and RW.
The system is programmed to result in either a protective or emergency stops. Emergency stop will be initiated when: 1. the emer-gency button is manually activated, 2. the light curtain is triggered, 3. intrusion in the RW and 4. mode-change button is not engaged after collaborative tasks. Protective stop will be initiated if the operator enters CW during automatic mode. After protective stop is initiated, the operator can go out of CW and engage the auto-continue button to restart robot motion. The robot controller can start a program with safe-signals from the master controller. In addition to setting and executing the manipulator poses, the robot controller initiates the gripping function via the dedicated vacuum controller[44].
6. Discussion
This section discusses aspects that describe a collaborative work-station with remarks on safety regarding large industrial robots. 6.1. Aspects of a Collaborative Workstation
Based on an analysis of the demonstrator presented in this article, three aspects of a collaborative workstations can be highlighted. They are: 1. Workspace, 2. Tasks and 3. Interaction. The layout of the de-monstrator can be described in terms of robot, operator and colla-borative workspaces. These workspaces together form the boundary of the workstation which according to ISO 12100 is the first step in
Fig. 8. Thefigure illustrates the role of the master safety controller that coordinates the functioning of various components in the laboratory demonstrator described infig: 5.
carrying out risk assessment. Within the collaborative workspace, a robot and operator(s) can carry out assembly tasks either independently or together (fig: 9).
The operator workspace can be understood as a physical location for: 1. an operator to monitor the system; 2. placement of equipment/ machinery to support operator task; 3. easily exit the plant in case of emergency and 4. plan and discuss activities with other co-workers. A robot workspace specifies the limit for the robotic system and asso-ciated equipment such as storage racks,fixtures, part handling equip-ment etc.
The demonstrator presented insection: 5is based on a case where the tasks are carried out manually. The manual tasks are delegated between the operator and robot. It may be noted that the division of labour implies additional tasks where, in addition to the assembly function, the operator should monitor the robotic system. The tasks in a collaborative workstation can be described in terms of operator, robot and collaborative tasks and are carried out within the corresponding workspaces. Robot tasks are a set of pre-programmed instructions car-ried out within the robot workspace. Whereas operator tasks include monitoring the workstation, preparing for upcoming assembly opera-tion etc., that are carried out in the operator workspace.
Interaction refers to the act of communication between two parti-cipants that intend to collaborate and in human-robot interaction this is enabled by interfaces (e.g., displays, gesture, buttons, and lamps). As discussed insection: 5, an operator has to interact with the system at different instances during one assembly cycle. The act of monitoring the system for fault, feedback from the system to begin collaborative task, communicating completion of collaborative tasks are examples of in-teraction.
6.2. Dynamic Workspaces
Krüger et al.[7]discusses a temporal characteristic of a collabora-tive system. That is, a collaboracollabora-tive workspace can be understood as a space and time sharing system. This temporal nature implies a time-frame within which a collaborative workspace can be occupied by ei-ther the robot or the operator. Insection 5, Collaborative Mode (CM) was used to describe the state of the system that permits operators to safely occupy the collaborative workspace. Designers of collaborative workstations needs to consider several challenged such as: 1. specify the state that permits start and end of CM, 2. specify the time-frame for which CM is safe and 3. specify a safe state for the machine when the parameters that specifies CM is no longer valid.
The mechanism for initiating collaborative and non-collaborative mode has been recognized as a particularly hazardous situation. ISO 10218-2 (sec: 5.11.4 in[22]) states that The change point between au-tonomous operation and collaborative operation is a particularly critical part
during risk assessment. As noted by Gopinath et al.[33], risk assessors can devote particular attention to evaluating probable designs based on foreseeable risks.
6.3. Risk Assessment
The risk assessment methodology is intended to support identi fica-tion and evaluafica-tion of potential hazards in order to achieve a safe and functional workstation. Therefore, risk analysis & evaluation are at the discretion of the assessment team and can result in varying solutions that are equally feasible. This phenomenon can be observed in the in-dustry where safety strategies vary and can be attributed to 1. local practices and understanding of safety, 2. Experience of the risk asses-sors. It can be motivated that experienced risk assessors are able to identify hazards that may seem trivial but has the potential to create serious injury.
A short summary of the hazards identified during the research phase was presented in table: 1. Prior to initiating collaborative mode (fig: 6c), the robot can move at high speed to synchronise with the moving line. If the operator were to accidentally enter the CW, the risk for serious injury are high. Therefore, a safeguarding strategy was se-lected where a laser scanner monitors the CW until collaborative mode is initiated (see No.1 intable: 1). To support mode awareness of the system, a warning lamp indicates the system-state to the operator. The operator can decide not to enter the CW if the lamp is red. The ha-zardous element is the fast-moving robot while the target is the op-erator. The mechanism that can initiate the hazard is the operator’s lack in awareness (sec: 2.3) of the system.
It may be noted that, in a human-populated environment, an op-erator can refer to any personnel who are in the vicinity to the work-station including the operator who is responsible for the workwork-station. Regardless of the exact safety solution, interviews with practitioners that tested the demonstrator suggest that the design of a collaborative system can prioritize intuitive and safe assembly operation. Therefore, a structured workspace (seesec: 6.1) with logical set of tasks supported by appropriate feedback devices is critical to achieve a safe and func-tional collaborative workstation.
7. Conclusion
This article presented a laboratory demonstrator that showcases a large industrial robot safely assisting an operator in assembly tasks on a continuously moving line. The demonstrator exemplifies three aspects, namely Workspaces, Tasks and Interaction that can be considered by system designers to plan and design a collaborative workstation.
The demonstrator also exemplifies a system architecture, consisting of safety-rated equipment and production machinery that mirrors cur-rent industrial practices. The integrated system showcases safety at a high technology readiness level and therefore demonstrates the feasi-bility of current safety technologies and standard practices for colla-borative robotics.
Additionally, the article motivates a Collaborative Mode of opera-tion to take into account the high risks associated when an industrial robot enters the Collaborative Workspace to carry out Collaborative Tasks. By explicitly recognizing Collaborative Mode, system designers can devote attention to: 1. specify parameters that define this mode and 2. make judgements on the hazards when collaborative tasks are being Fig. 9. Illustration of operator, robot and collaborative workspaces which
to-gether form the boundary of a robotic system where an operator and a robot collaborative to complete manufacturing operations.
Supervision, Funding acquisition.Micael Derelöv: Conceptualization, Supervision, Resources. Åke Gustafsson: Conceptualization, Methodology, Investigation, Visualization, Supervision, Project ad-ministration. Stefan Axelsson: Conceptualization, Methodology, Investigation, Visualization, Supervision, Project administration. Declaration of Competing Interest
The authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manu-script.
Acknowlgedgments
This work has been primarily funded under the FFI program by Vinnova and the authors would like to graciously thank them for their support. We would like to thank the members of the research project Collaborative Team of Man and Machine for their valuable input and suggestions.
References
[1] T. Brogȧrdh, Present and future robot control development – An industrial per-spective, Annual Reviews in Control 31 (1) (2007) 69–79.
[2] B. Siciliano, O. Khatib, Springer Handbook of Robotics, Springer-Verlag New York, Inc., Secaucus, NJ, USA, 2007.
[3] Swedish Standards InstituteStockholm, Sweden, SS-ISO 10218-1:2011– Robots and robotic devices– Safety requirements for industrial robots – Part 1: Robot), (2011). [4] F. Vicentini, Terminology in safety of collaborative robotics, Robotics and
Computer-Integrated Manufacturing 63 (2020) 101921,https://doi.org/10.1016/j. rcim.2019.101921.
[5] The International Organization for Standardization, ISO/TS 15066– Robots and robotic devices– Collaborative robots, (2016).
[6] G. Reinhart, J. Werner, F. Lange, Robot based system for the automation offlow assembly lines, Production Engineering 3 (1) (2009) 121–126,https://doi.org/10. 1007/s11740-008-0143-z.
[7] J. Krüger, T. Lien, A. Verl, Cooperation of human and machines in assembly lines, CIRP Annals - Manufacturing Technology 58 (2) (2009) 628–646.
[8] N. Pedrocchi, F. Vicentini, M. Matteo, L. Molinari, Safe Human-Robot Cooperation in an Industrial Environment, International Journal of Advanced Robotic Systems 10 (27) (2013) 13.
[9] J. Krüger, R. Bernhardt, D. Surdilovic, Intelligent Assist Systems for Flexible Assembly, CIRP Annals - Manufacturing Technology 55 (1) (2006) 29–32. [10] F. Ore, L. Hanson, N. Delfs, M. Wiktorsson, Human industrial robot collaboration–
development and application of simulation software, International Journal of Human Factors Modelling and Simulation 5 (2) (2016) 164–185.
[11] A. Cherubini, R. Passama, A. Crosnier, A. Lasnier, P. Fraisse, Collaborative manu-facturing with physical human-robot interaction, Robotics and Computer-Integrated Manufacturing 40 (2016) 1–13,https://doi.org/10.1016/j.rcim.2015. 12.007.
[12] J.A. Marvel, Performance metrics of speed and separation monitoring in shared workspaces, IEEE Transactions on Automation Science and Engineering 10 (2)
[19] Kuka AG,https://www.kuka.com/(site accessed– 11 December).
[20] C.A. Ericson, Hazard analysis techniques for system safety, John Wiley & Sons, 2015.
[21] J. Etherton, M. Taubitz, H. Raafat, J. Russell, C. Roudebush, Machinery risk as-sessment for risk reduction, Human and Ecological Risk Asas-sessment 7 (7) (2001) 1787–1799,https://doi.org/10.1080/20018091095393.
[22] Swedish Standards InstituteStockholm, Sweden, SS-ISO 10218-2:2011– Robots and robotic devices– Safety requirements for industrial robots – Part 2: Robot systems and integration), (2011).
[23] V. Gopinath, K. Johansen, Å. Gustafsson, S. Axelsson, Collaborative assembly on a continuously moving line– an automotive case study, Procedia Manufacturing 17 (2018) 985–992,https://doi.org/10.1016/j.promfg.2018.10.105. 28th International Conference on Flexible Automation and Intelligent Manufacturing (FAIM2018), June 11–14, 2018, Columbus, OH, USA
[24] W. Chow, Assembly Line Design: Methodology and Applications, Marcel Dekker, Inc., 1990.
[25] H. Geng, Manufacturing Engineering Handbook, Second Edition, McGraw-Hill Professional, 2016.
[26] B. Diffner, Combining Flexibility and Efficiency in Automotive Assembly : Preparing for New Powertrain Vehicles. Linköping : Department of Management and Engineering, Linkping University, 2011.
[27] Swedish Standards InstituteStockholm, Sweden, SS-ISO 12100:2010– Safety of Machinery– General principles of Design – Risk assessment and risk reduction (ISO 12100:2010), (2010). SS-EN ISO 12100:2010
[28] N.G. Leveson, Engineering a Safer World: Systems Thinking Applied to Safety, Engineering systems, MIT Press, 2011.
[29] The International Electrotechnical Commision, IEC TS 62046:2008– Safety of machinery - Application of protective equipment to detect the presence of persons, (2008).
[30] N.B. Sarter, D.D. Woods, How in the World Did We Ever Get into That Mode? Mode Error and Awareness in Supervisory Control, Human Factors: The Journal of the Human Factors and Ergonomics Society 37 (1) (1995) 5–19,https://doi.org/10. 1518/001872095779049516.
[31] T.B. Sheridan, R. Parasuraman, Human-Automation Interaction, Reviews of Human Factors and Ergonomics 1 (1) (2005) 89–129,https://doi.org/10.1518/ 155723405783703082.
[32] C. Upton, G. Doherty, Extending Ecological Interface Design principles: A manu-facturing case study, International Journal of Human Computer Studies 66 (4) (2008) 271–286.
[33] V. Gopinath, K. Johansen, Understanding situational and mode awareness for safe human-robot collaboration: case studies on assembly applications, Production Engineering 13 (1) (2019) 1–9,https://doi.org/10.1007/s11740-018-0868-2. [34] V. Gopinath, M. Derelöv, J. Ölvander, K. Johansen, Development of Demonstrators
for Industrial Safety An Interactive Perspective, International Journal of Occupational Safety and Ergonomics (JOSE) Under Review– JOSE-2019-0377 (2020).
[35] V. Gopinath, On Safe Collaborative Assembly With Large Industrial Robots, Linköping University, Machine Design, Faculty of Science and Engineering, (2019),
https://doi.org/10.3384/diss.diva-162134.
[36] R. Yin, Case Study Research: Design and Methods, Applied Social Research Methods, SAGE Publications, 2014.
[37] P.D. Leedy, J.E. Ormrod, Practical Research: Planning and Design, Pearson Education Limited, 2013.
[38] M. Jonsson, On Manufacturing Technology as an Enabler of Flexibility : Affordable Reconfigurable Tooling and Force-Controlled Robotics. Linköping Studies in Science and Technology. Dissertations: 1501, (2013).
[39] A. Björnsson, Automated layup and forming of prepreg laminates, Linköping UniversityLinköping University, Manufacturing Engineering, Faculty of Science and Engineering, (2017),https://doi.org/10.3384/diss.diva-137488.
[40] L.T.M. Blessing, A. Chakrabarti, DRM, a Design Research Methodology, Springer Publishing Company, Incorporated, 2009. edition 1st
[41] V. Gopinath, K. Johansen, A. Gustafsson, Design criteria for conceptual end effector for physical human robot produciton cell, Swedish Production Symposium, (2014). [42] V. Gopinath, Collaborative assembly on a continuously-moving line, (2019). site
accessed– 09-October
[43] Festo AG, (2019). https://www.festo.com/hq/en/cms/festo.htm , site accessed– March 2019
