Integrating Bio Principles in Assembly Task Design to Enhance Symbiotic Human-Robot Collaboration

doi:10.21203/rs.3.rs-3886296/v1

Download PDF

Research Article

Integrating Bio Principles in Assembly Task Design to Enhance Symbiotic Human-Robot Collaboration

https://doi.org/10.21203/rs.3.rs-3886296/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The advancement of Human-Robot Collaboration (HRC) in industrial environments has underscored the importance of establishing harmonious and symbiotic relationships between humans and robots. This study, in line with the principles of Industry 5.0, proposes an approach to support the integration of human capabilities with advanced robotics, enhancing collaborative productivity and fostering a paradigm shift towards a more interactive and beneficial human-robot symbiosis.

Prior research has established the basic principles of Symbiotic Human-Robot Collaboration (SHRC), but has often neglected the critical problem of how to conduct collaborative tasks to exploit the potential of these symbiotic interactions.

This paper presents a novel methodology to support the design of protocols for collaborative tasks, with the aim of promoting positive symbiotic interactions between humans and collaborative robots. The focus is on developing tasks that naturally integrate positive symbiotic interactions, which involves determining task performers and optimizing the mutual benefits derived from task execution.

A case study is presented to illustrate the practical application of this methodology in a real-world context.

Human-Robot Collaboration

Symbiotic Human-Robot Collaboration

Collaborative Robotics

Human-Robot Symbiosis

Collaborative Assembly

Task allocation

Human-Centric Manufacturing

Industry 5.0

The rapidly evolving field of Human-Robot Collaboration (HRC) has reached a pivotal juncture where the synthesis of human and robotic capabilities can redefine collaborative endeavors (Inkulu et al., 2022; Lu et al., 2022). This intersection of humans and robots has unlocked unprecedented possibilities for augmenting human potential with robotic precision, adaptability, and efficiency (Faccio et al., 2023). As technological advancements continue to blur the lines between human and machine, it becomes increasingly important to explore and harness the potential of these collaborative relationships (Barravecchia et al., 2023b).

The essence of HRC lies in orchestrating harmonious and positive symbiotic relationships between human operators and robots (Simões et al., 2022). A positive symbiotic relationship, in this context, indicates an interaction where both human and robot benefit from each other’s distinct capabilities, thereby contributing to the overall efficiency, productivity, and safety of the collaborative tasks (Barravecchia et al., 2023a). While the importance of such a symbiotic interaction is acknowledged, its practical realization presents several challenges and opportunities (Gervasi et al., 2020; Li et al., 2021).

To realize an effective symbiotic relationship, cobots are increasingly being equipped with advanced technologies that significantly expand their capabilities (Barravecchia et al., 2023). The integration of these technologies results in collaborative systems that elevate cobots beyond simple physical tasks. In this study, the term collaborative systems is defined in a broad sense, extending beyond the traditional emphasis on physical machinery, such as cobots. As defined in this context, a collaborative system comprises a range of equipment designed to facilitate and support human activities, thereby providing an infrastructure for HRC (Barravecchia et al., 2023). The design of these systems is focused on augmenting human capabilities, ensuring that the collaborative effort effectively synergizes the strengths of both humans and technology.

Equipped with dynamic communication features, intuitive interactions, experiential learning and problem solving, such systems can represent a significant advancement in the potential of collaborative robotics (Wang et al., 2017).

This article explores the evolving landscape of Symbiotic Human-Robot Collaboration (SHRC) within manufacturing contexts, with a particular focus on its role in enhancing assembly processes. Existing research in this domain has largely revolved around the technological problems of establishing SHRC (Barravecchia et al., 2023; Gervasi et al., 2022b; Verna et al., 2023; Wang et al., 2019). This prevalent orientation has overshadowed the crucial aspect of how assembly tasks should be allocated and be collaboratively executed to optimize the symbiotic exchanges between the human and the robot (Barravecchia et al., 2023a). This gap in the existing literature underscores the need for a methodological framework that delves into the protocols of task execution, aiming to maximize the positive symbiotic exchanges in a HRC setting.

Addressing a significant research question, this article seeks to determine: how can collaborative task be designed to facilitate positive symbiotic interactions between humans and collaborative systems in assembly processes? A novel method is proposed to provide practical solutions for the allocation of collaborative tasks and the development of corresponding task protocols. This paper adopts the term task protocol to describe the series of specific actions and interactions that both the human operator and the collaborative system are expected to perform and engage in during their collaborative efforts. This concept encapsulates the individual tasks assigned to each agent, the dynamics of their interplay, communication methods, and the sequential flow of activities. The purpose of defining the task protocol for the assembly process is to provide a clear and structured framework to guide the collaborative activities, ensuring an effective partnership and, consequently, a positive symbiotic relationship.

The proposed method is based on the idea that promoting positive symbiotic interactions between humans and robots, especially in assembly tasks, is essential to fully harness the potential of HRC (Faccio and Cohen, 2023). The purpose is move towards a comprehensive framework that captures the essential elements of 'Who' performs and 'How' tasks are carried out in a joint effort, aiming to improve symbiotic interactions.

The remainder of the paper is structured as follows. Section 2 presents a conceptual background on SHRC, offering a foundational understanding of the interplay between humans and robots in collaborative settings. Section 3 provides an overview of previous research on collaborative task allocation, examining existing approaches and their implications. In Section 4, the focus shifts to collaborative task design, where insights into the design aspects of collaborative tasks are discussed. Section 5 introduces the proposed methodology for the design of symbiotic task for collaborative assembly, detailing the steps and considerations involved in this approach. An applied case study is presented in Section 6, illustrating the practical implementation of the methodology in a real-world scenario. Finally, the concluding section synthesizes the key findings, explores the broader implications of the research, and proposes paths for future exploration in this field.

Before considering the design of symbiotic tasks, it is essential to deepen the understanding of the concept of Symbiotic Human-Robot Collaboration (SHRC). SHRC refers to the interaction between humans and robots in a cooperative and mutually beneficial manner, where both agents work together to achieve a common goal and each enhances the capabilities of the other (Wang et al., 2017). Wang et al. (2019) outlined six key requirements of SHRC systems:

Autonomy: leadership and roles are taken on and changed during the process with the aim of exploiting the skills of both agents.
Context-awareness: agents make decisions and take actions based on the current state of the process and the conditions of the working environment.
Communication: the symbiosis requires constant interaction between human and robot, necessitating multimodal and bi-directional communication.
Digital Twin: agents use (at least in part) shared representations of the environment in which they operate as a prerequisite for coordinating their shared goals, roles, plans and activities.
Learn: the system enables learning and adaptation to changing working conditions by providing continuous feedback to the agents involved in the process.
Safety: the safety of the human operator is guaranteed in the shared environment, even under unpredictable circumstances.

Understanding the essentials of SHRC lays the groundwork for appreciating the diverse nature of interactions within HRC, particularly in assembly processes. Barravecchia et al. (2023a) elaborate on this by exploring various symbiotic human-robot relationships, framing them in a scheme drawn from the natural symbiotic context, as described below (see Fig. 1):

Mutualism: This occurs when collaboration between humans and robots results in mutual benefits. For instance, robots taking over repetitive tasks enhance overall efficiency, allowing humans to focus more on aspects like quality control and strategic decision-making.
Commensalism: In these scenarios, one party benefits while the other is neither harmed nor helped significantly. A prime example is a cobot lifting heavy loads. This reduces the physical effort and risk for human operators, enhancing their workplace safety, while the robot's functionality remains unaffected.
Parasitism: Here, one party benefits at the expense of the other. An example in HRC is when a robot's operating speed exceeds the comfortable pace of its human counterpart. This mismatch can lead to increased stress for the human worker, a higher likelihood of errors, and decreased job satisfaction, all of which negatively impact the overall efficiency of the task.
Amensalism: This type of interaction occurs when the activity of one collaborative agent hampers the other, without any discernible benefit to itself. An example could be a scenario where the presence of a robot in a shared workspace restricts the movement or efficiency of the human worker, without improving the robot's own performance.
Incompatibility: This represents situations where both parties are negatively impacted by the collaboration. This could manifest in an industrial setting where the integration of a robot into a human team leads to disruptions in workflow, misunderstandings in task execution, or increased complexity in the process, thus reducing efficiency and productivity for both.
Neutralism: These are relations where the collaboration between human and robot has negligible impact on either party. This might occur in scenarios where the tasks performed by humans and robots are so distinct and independent that their collaborative interaction doesn’t meaningfully change the output or process.

This taxonomy can be a useful reference point for designing and analyzing symbiotic relationships (Barravecchia et al., 2023a).

To fully realize and enhance SHRC, it is critical to comprehend the symbiotic factors that influence the relationship.Barravecchia et al. (2023a) identified three distinct symbiotic factors (action, guidance and protection) and their corresponding dimensions relevant to HRC.

The action factor involves the exchange of concrete actions required for task completion. For example, in an assembly process, a robot may position parts while a human operator secures bolts, exemplifying an integral exchange of actions for task completion. Two components are related to this factor:

Effort: agents’ contribution to the effort required to complete the task.
Speed: agents’ influence on the progression of the task.

The Guidance factor is the ability of one agent to instruct the other to complete a task. This is the case of a collaborative system guiding a human operator through assembly operations by providing real-time feedback and guidance to facilitate effective task completion. The components of the guidance factor are:

Knowledge: agents’ knowledge needed to complete the task.
Decision-making: agents’ inferential ability to determine the task to be performed and the method of completion.

Protection, the third factor, is concerned with protecting the other collaborative agent from potential threats arising from the collaboration. An illustrative scenario is where a robot performs repetitive or physically demanding tasks, reducing the cognitive and physical load on the human operator. The components of the protection factor are:

Ergonomics: agents’ ability to alter the ergonomic well-being of the other, affecting physical and psychological comfort.
Safety: agents’ ability to either expose or protect the other agent from potential risks or threats.

Gaining insight into these symbiotic factors is crucial for refining interactions in SHRC, providing a structured framework for enhancing synergy in various collaborative scenarios.

In the field of HRC, task allocation has emerged as a pivotal area of study (Kiyokawa et al., 2023). The primary objective of task allocation approaches is to distribute assembly tasks between human workers and cobots (Ranz et al., 2017).

The literature primarily categorizes task allocation into two main distinct paradigms: static task allocation and dynamic task allocation (Petzoldt et al., 2023).

In static task allocation, roles are determined based on an analysis that compares the suitability of both human and robotic agents for specific tasks. Once set, these roles remain fixed throughout the assembly process. Most static allocation methods evaluate suitability by matching the abilities of both humans and robots to the characteristics of the components being assembled and the demands of the related assembly task (Müller et al., 2016). Advanced methodologies delve deeper into the suitability assessment by incorporating additional factors like ergonomics, task completion durations, and cycle times. Other strategies promote a wider view of static task allocation, shifting from a task-focused approach to considering the entire assembly process. Such methods employ algorithms to formulate task allocation schedules, aiming to enhance the overall efficiency and effectiveness of the assembly (Lee et al., 2022; Liau and Ryu, 2022). A notable extension to traditional static task allocation methods involves the use of simulation environments (Bänziger et al., 2020). One primary utility of simulation environments is their ability to compare various task allocation schedules (Michalos et al., 2018; Tsarouchi et al., 2017). By visualizing different alternatives in a simulated assembly process, it is possible to systematically evaluate the impact of each allocation strategy on the overall assembly performance. Furthermore, simulation environments can be integrated with algorithms designed for multi-criteria optimization (Ranz et al., 2017). These algorithms consider a set of factors including efficiency, effectiveness, ergonomics, and cycle times to optimize the allocation of tasks between human workers and robots.

Conversely, in dynamic task allocation, emphasis is placed on the adaptability of assignments and real-time responsiveness, allowing for a flexible and responsive approach to the changing demands of the assembly process. Within this framework, there are two primary approaches that are commonly identified in the literature: (i) reactive task allocation and (ii) proactive task allocation (Petzoldt et al., 2023).

The reactive task allocation approach is defined by a centralized allocation plan that is subject to reactive adjustments in response to real-time changes or disturbances that may occur during the assembly process, such as unexpected collisions or sudden changes in the workload of the human operator (Antonelli and Bruno, 2019; Pupa and Secchi, 2021) .This approach may involve the dynamic adaptation of individual task assignments (Tsarouchi et al., 2016) or, in more extreme cases, a complete re-planning of all remaining tasks in the assembly process (Cesta et al., 2018). This approach is flexible and responsive, allowing for adjustments to be made on-the-fly in order to accommodate unforeseen circumstances of the assembly process. The proactive task allocation approach takes a different stance by placing the human operator at the forefront of the assembly process. In this approach, the human operator is empowered with the autonomy to select the tasks they wish to perform (Liu et al., 2019). Based on the human's actions, the system then dynamically decides which tasks can be performed by the robot. The objective here is to identify tasks that can be performed in parallel with the human operator's tasks, serving as preparatory work for subsequent steps in the assembly process (Fiore et al., 2016). This approach can enhances the efficiency of the assembly process and ensures that the human operator is able to work at their own pace, with the robot providing support as needed.

Table 1 proposes a comparative analysis of Static Task Allocation and Dynamic Task Allocation, highlighting the advantages and disadvantages of each approach in terms of flexibility, responsiveness, complexity, human operator involvement, and efficiency.

Table 1

*Comparative analysis of Static Task Allocation and Dynamic Task Allocation approaches in Human-Robot Collaboration* (Petzoldt et al., 2023).
CHARACTERISTIC	STATIC TASK ALLOCATION	DYNAMIC TASK ALLOCATION
FLEXIBILITY	Limited; roles and tasks are pre-defined and remain fixed.	High; tasks and roles can be adapted in real-time to meet the needs of the assembly process.
RESPONSIVENESS	Limited; lacks the ability to respond to changes and disturbances in real-time.	High; can quickly adjust to real-time changes and disturbances in the assembly process.
COMPLEXITY	Moderate; methodologies range from simple to complex, depending on factors considered during the allocation process.	High; requires real-time decision-making and coordination between human operators and robots.
HUMAN OPERATOR INVOLVEMENT	Moderate; human operators have clearly defined roles and tasks.	High; human operators can make autonomous choices and have more control over the assembly process.
EFFICIENCY	High; tasks are allocated based on optimal suitability, leading to an efficient assembly process.	Moderate; efficiency may be affected by the need to make real-time adjustments.

Both Static Task Allocation and Dynamic Task Allocation are valuable approaches in the field of HRC. Static Task Allocation, with its fixed roles and tasks, offers a high degree of predictability and efficiency. The tasks are allocated based on a detailed analysis of optimal suitability, ensuring that the strengths of both human operators and robots are leveraged in the most effective manner. This approach is particularly advantageous in environments where tasks and roles remain constant, allowing for a streamlined and efficient assembly process.

On the opposite, Dynamic Task Allocation is more flexible and responsive, with the ability to adapt to real-time changes in the assembly process. This approach is well-suited for unpredictable environments that require real-time adjustments to meet the demands of the assembly process.

The exploration of collaborative task design within assembly processes is relatively unexplored, with a handful of recent studies providing some preliminary insights into the interplay between humans and robots.

The work of Mulesa et al. (2023) shed light on the project planning phase by emphasizing the construction of a task hierarchy and its subsequent distribution among executors. Their approach is marked by the introduction of efficiency indicators and the proposition of iterative methods aimed at evaluating the minimal duration and cost associated with the task.

The study by Garcia et al. (2023) introduces a novel framework for HRC, leveraging deep learning models to manage assembly processes conducted either individually or collaboratively by humans and robots.

Further contributing to the ergonomic aspect of collaborative task design, the research conducted by Navas-Reascos et al. (2022) presents a prototype and simulation to integrate a collaborative system in a wire harness assembly process. This new emphasis on ergonomics highlights the importance of considering human well-being in the design of collaborative tasks (Gervasi et al., 2022a).

Dmytriyev et al. (2022) focus on a cooperative assembly task, using a Brain-Computer Interface (BCI) to relay commands to the collaborative robot. This allows the operator to switch between independent and cooperative assistance modes seamlessly. The study highlights the potential of integrating advanced technologies to enhance communication and cooperation.

Moreover, the integration of heterogeneous models for safety-critical mechatronic systems, as explored by Mhenni et al. (2022), offers insights into the application of Model-Based Systems Engineering (MBSE), Model-Based Safety Assessment (MBSA), and multi-physics modeling in designing collaborative workplaces for aircraft assembly. This integrative approach underscores the multifaceted nature of collaborative task design, necessitating a convergence of diverse disciplines and domains to address the requirements and considerations inherent in collaborative assembly processes.

Lastly, the work of Huang et al. (2021) introduces a protocol linking design and construction intent with algorithmic planning for robotic assembly of complex structures. This protocol enables designers to articulate their design intentions and process knowledge, providing a structured framework to define complete sequence of robot motions.

A critical examination of the literature reveals the absence of structured methods and a predominant focus on specific case studies, limiting the generalizability of the findings. The current body of work predominantly centers on isolated aspects of collaboration such as ergonomics or communication, thereby lacking a comprehensive approach that encompasses the multifaceted nature of collaborative task execution.

The need for a structured method able to consider all dimensions of a collaborative task design is evident. Such a method should not only integrate the diverse aspects of collaboration but also facilitate the development of generalized frameworks that can be applied across different scenarios and contexts.

This Section presents a novel method to assist the design of symbiotic task for collaborative assembly. The proposed method aims to guide the definition of the task protocol. To achieve this goal, the following steps are considered (see Fig. 2):

The initial step involves characterize the activities to be performed in the assembly process. In cases where the process is intricate, it is impractical to dissect individual activities in isolation. In such cases, categorizing activities into specific task categories can be an effective method to simplify the problem domain (Lihui Wang and Feng, 2011). This step, referred to as task categorization, is designed to address the question, "What tasks are required to complete the assembly?"

The second step focuses on assigning the identified task categories to the appropriate collaborative agents, whether human or collaborative systems. This step, referred to as task allocation, is dedicated to answering the question, “Who is responsible for executing the identified tasks?”

The third step involves detailing the actions of the collaborative agents. After establishing the tasks and assigning responsibility, the final step involves outlining the behaviors and interactive patterns of the collaborative agents. This step, named Task Protocol Generation produces the formulation of the complete task protocol, thereby providing an answer to the question, "How should the task be executed?"

The primary objective of the proposed approach is to design collaborative tasks that maximize symbiotic exchanges between humans and collaborative systems, ensuring that they complement each other's strengths and mitigate each other's weaknesses. The following sub-section provides an overview of the proposed approach. Subsequently, a case study will be presented to exemplify the practical application of the method.

5.1 Input of the method

The proposed methodology requires two primary inputs:

Bill of Materials (BOM): The Bill of Materials is a comprehensive list that outlines all the components, materials, and sub-assemblies required to manufacture the final product (Reid and Sanders, 2020). It is a reference point for understanding the various elements involved in the assembly process and their relationships to each other.
List of Ordered tasks: The second input is a detailed and ordered list of tasks to complete the assembly process. This list should outline each step involved in the assembly, from the initial identification and handling of components to the final checking and adjustment of the assembled product.

Step 1: Task categorization

An essential step for organizing the various tasks involved in an assembly process is the initial phase of task categorization. This categorization needs to take into account two aspects.

The first is represented by the elementary task type. The wide variety of tasks that constitute an assembly process can be reduced to a limited number of elementary task types with common execution attributes. In detail, five elementary task types can be considered as fundamental components of the assembly process (Boothroyd et al., 2010; Radkowski et al., 2015):

Identification and Handling: this involves recognizing the necessary components and their handling.
Alignment: this task requires aligning the components in the correct orientation and position.
Joining: this involves connecting the components together to form the final product or subassembly.
Checking: this task involves inspecting the assembly to ensure that it meets the required specifications.
Adjustment: this involves making any necessary adjustments to the assembly to ensure a correct functioning.

The second aspect that can be considered for categorizing a task involves the inherent properties of the component involved. This element plays a crucial role in selecting the most appropriate agent for task completion, since tasks of the same type can differ substantially depending on the components involved. In light of this, by identifying the specific characteristics of each component, it is possible to allocate different tasks accordingly (Malik and Bilberg, 2019). The following are potential factors that can significantly influence task execution and can be used to classify components:

Grippability: The ease of gripping and manipulating a component during the assembly process is crucial in determining who will perform the assembly task and how it will be done (Malik and Bilberg, 2019). Certain types of components may pose challenges for humans or collaborative systems in terms of grippability. For instance, some components may be too heavy for human workers to handle and manipulate, while collaborative systems may lack the necessary grippers or end-effectors to handle certain types of components.
Feeding: It basically refers to the ease with which the component can be located and retrieved from the feeding zone (Malik and Bilberg, 2019).
Orientation: This refers to the specific position or alignment that a component must have in relation to other components or the assembled product as a whole. Some components may require a specific orientation that challenges the capability of humans or collaborative systems to achieve (Malik and Bilberg, 2019) .
Temperature: The temperature of a component during assembly can affect how easily it can be handled and assembled (ISO, 2008). In contexts where components are exposed to extreme temperatures, distinguishing the temperature sensitivity of components can be essential to designing the assembly process.
Surface treatment: The presence of paints, coatings, lubricants, etc., can affect the task design choices. In cases where components have surface treatments that influence assembly activities, it is important to distinguish them during this task categorization phase(Borboni et al., 2023).
ESD Sensitivity: Electrostatic discharge sensitivity is a critical factor in some assembly processes, particularly in the electronics industry. If some components have specific ESD requirements, this should be distinguished (Tamminen et al., 2016).
Cost: Components with different costs may require different assembly activities (e.g., inspections, handling, etc.). It is possible to classify components based on this aspect as well (Barravecchia et al., 2023b; Huang and Shiau, 2009).

The list above includes some common criteria, but specific contexts may require additional or alternative classification criteria for components.

Criteria can be used individually (e.g., grippable/ungrippable component) or combined (e.g., grippable and inexpensive component, grippable and expensive component, ungrippable and inexpensive component, ungrippable and expensive component).

This phase combines elementary task types with component properties to identify detailed task categories. These categories provide a clear understanding of the assembly process, crucial for facilitating the subsequent task design phases.

Step 2: Symbiotic task allocation

After defining and categorizing the tasks, the next step is to assign them to the appropriate agent, whether it be a human worker or a collaborative system. As previously introduced, the goal is designing the task to enhance the symbiosis between the human operator and collaborative system, thereby maximizing exchanges on the six components of the SHRC factors outlined in Section 2: effort, speed, knowledge, decision-making, ergonomics, and safety.

The key focus of this approach is to determine, for each individual task, which of the two agents – human or collaborative system – will assume the role of the leader and supporter. The leader assumes the primary responsibility for completing the task, and the supporter assists and helps the leader in completing the task.

To facilitate the decision-making process, the proposed methodology involves assessing the capabilities of both humans and collaborative systems in relation to each component of SHRC. This evaluation should take into consideration the unique strengths and weaknesses of both agents in relation to the specific task category at hand.

To assist the experts in making informed decisions, Table 2 provides a series of questions that are designed to assess the capabilities of humans and collaborative systems across the six symbiotic components. Each question aims to determine which agent is better equipped to handle the task category with respect to the specific dimension.

Table 2

Questions to assess the capabilities of humans and collaborative systems across symbiotic components.
SYMBIOTIC COMPONENT	QUESTION	DESCRIPTION	IMPLICATIONS
Effort	Who performs better in terms of effort required for the task?	Assessing the physical or cognitive load required to perform the task.	A lower effort suggests a more efficient process.
Speed	Who performs better in terms of speed and efficiency for the task?	Evaluating the time taken to complete the task.	Faster completion typically results in higher productivity.
Knowledge	Who is better equipped with the necessary knowledge and expertise for the task?	Determining the level of technical or domain-specific expertise required for the task.	The agent with the relevant knowledge is likely to perform the task more accurately.
Decision-Making	Who is better equipped to make informed and accurate decisions for the task?	Assessing the capability to process information and make the right decisions based on the task requirements.	A good decision-making process can impact the task outcome.
Ergonomics	Who provides a more ergonomic solution for the task?	Evaluating the comfort and ease of performing the task.	An ergonomic solution can reduce physical and psychological strain and enhance productivity.
Safety	Who ensures a safer execution of the task?	Assessing the potential risks involved and the measures taken to mitigate those risks.	Ensuring safety is critical in any task.

For each symbiotic component responses should be provided in the form of “Human”, “Collaborative System”, “Equal”, or “Not Applicable”. "Human" indicates that the human agent is better equipped to handle the task category in that specific dimension; "Collaborative System" suggests that the robotic system is more suitable; "Equal" implies that both agents are equally able; "Not Applicable" indicates that the dimension in question is not relevant or cannot be assessed for the particular task category at hand.

A straightforward method for task allocation, based on these criteria, suggests that if the majority of evaluated dimensions indicate one agent as more suitable, then that agent should be designated the responsibility for executing the task category. In cases where there is an equal suitability between agents, the task category should be delegated to the collaborative system.

As alternatives to this basic approach, more sophisticated decision-making mechanisms can be applied (Cinelli et al., 2020; De Montis et al., 2004; Ishizaka and Nemery, 2013). For instance, assigning different weights to the symbiotic components or considering veto conditions. This would allow the evaluation process to be tailored to the needs of specific applications.

This methodology is classified as a static task allocation approach, allowing for clear and structured role allocation based on pre-defined criteria.

Step 3: Task protocol generation

Once "what" needs to be done and "who" will be responsible are defined, the next step is to establish "how" the collaborative tasks should be executed. This entails the development of a task protocol. Each task category, identified during the task categorization phase and allocated in the task allocation phase, is mapped with the activities of the leader and the supporter.

Table 3 and Table 4 can provide assistance in this process. They provide a comprehensive summary of the basic activities for the leader (whether human or robot) and the supporter (whether human or robot) for each task type, with the aim of optimizing the symbiotic exchanges between humans and robots and ensuring effective collaboration. Detailed descriptions of the support activities listed in the Tables A.1 and A.2 provided in Appendix A.

The supporting activities were identified through a process aimed at strengthening SHRC. This included:

Capability analysis: Assessing the capabilities and needs of both human operators and collaborative systems.
Ergonomic assessments: Ensuring the safety and comfort of human operators.
Field observations and operator feedback: Identifying where system support is most effective.
Industry benchmarking and expert advice: Incorporating best practice and guidance into the design of support activities.

This approach developed activities that can be adapted to various assembly contexts. The deliberately generic activity descriptions (see Tables A.1 and A.2 provided in Appendix A) allow for flexibility in application to different technologies. For example, the “identification of a target component” (see Table A.1, first row) can be achieved through various technologies and procedures, such as using a screen, a laser, verbal description of component characteristics, or simply physically pointing to it. This flexibility ensures that the guidelines remain relevant and adaptable, maintaining their utility in the ever-changing landscape of collaborative assembly.

Table 3

Summary of the activities of an assembly process, with the human as the leader and the collaborative system acting as the supporter. Note: the dots in the last six columns represent the dimensions in which the support activities can significantly enhance human-robot symbiosis. Table A.1 in Appendix A details the supportive activities of the collaborative system.
	HUMAN LEADER ACTIVITY		SYMBIOTIC FACTORS
			Action		Guidance		Protection
			Effort	Speed	Knowledge	Decision Making	Ergonomics	Safety
TASK TYPE		COLLABORATIVE SYSTEM SUPPORTIVE ACTIVITY	Effort	Speed	Knowledge	Decision Making	Ergonomics	Safety
Identification & Handling	The human operator identifies the correct component from the feed area and moves the selected component to the work area.	Target component		●	●		●
		Real-time identification feedback		●	●		●
		Sorting assistance		●	●		●
		List of activity sharing		●	●	●	●
		Component defects detection		●		●	●
		Lifting assistance	●	●			●
		Component pre-positioning		●	●		●
		Posture monitoring	●				●
		Real-time handling feedback		●	●		●	●
		Grip assistance		●	●		●	●
		Safety alerts						●
		Stress-driven handling scheduling				●	●
		Process-driven handling scheduling		●		●
		Tool handover	●	●	●		●
		Data logging			●	●		●
Alignment	The human operator aligns and places the component in its designated mounting location.	Target component position		●	●		●
		Active alignment stabilization	●	●			●
		Subassembly pre-orientation	●	●	●		●
		Operator-optimized subassembly orientation	●	●			●
		Real-time alignment feedback (profile matching)		●	●		●
		Alignment sequence sharing	●	●	●
		Stress-driven alignment scheduling					●
		Process-driven alignment scheduling	●			●
		Safety alerts			●			●
		Precision positioning		●	●		●
		Alignment instructions	●	●	●
		Visual field optimization		●			●
Joining	The human operator physically joins the components.	Synchronized component joining	●	●	●		●
		Secure holding	●	●			●
		Subassembly pre-orientation		●	●
		Operator-optimized subassembly orientation	●	●			●
		Joining instructions	●	●	●
		Real time joining feedback		●	●	●	●
		Ergonomic subassembly positioning					●
		Adaptive ergonomic subassembly positioning					●
		Stress-driven joining scheduling					●
		Process-driven joining scheduling	●			●
		Safety alerts						●
		Tool handover	●	●	●		●
		Visual field optimization		●			●
		Data logging			●	●		●
Checking	The human operator performs a quality assessment to confirm that the assembly complies with specifications.	Human-error mitigation		●	●		●
		Operator-optimized product orientation	●	●			●
		Inspection instructions	●	●	●
		Safety alerts						●
		Ergonomic product positioning					●
		Adaptive ergonomic product positioning					●
		Stress-driven checking scheduling					●
		Process-driven checking scheduling	●			●
		Defect tracking assistance		●	●
		Visual field optimization		●			●
		Tool handover	●	●	●		●
Adjustment	The human operator rectifies defects in the assembled product to ensure that the final product meets specifications.	In-process validation	●	●		●
		Real-time adjustment guidance		●	●		●
		Operator-optimized product orientation	●	●			●
		Adjustment instructions	●	●	●
		Safety alerts						●
		Ergonomic product positioning					●
		Adaptive ergonomic product positioning					●
		Stress-driven adjustment scheduling					●
		Process-driven adjustment scheduling	●			●
		Visual field optimization		●			●
		Tool handover	●	●	●		●

Table 4

Summary of the activities of an assembly process, with the collaborative system as the leader and the human acting as the supporter. Note: the dots in the last six columns represent the dimensions in which the support activities can significantly enhance human-robot symbiosis. Table A.2 in Appendix A details the supportive activities of the human operator.
			SYMBIOTIC FACTORS
			Action		Guidance		Protection
			Effort	Speed	Knowledge	Decision Making	Ergonomics	Safety
TASK TYPE	COLLABORATIVE SYSTEM LEADER ACTIVITY	HUMAN SUPPORTIVE ACTIVITY	Effort	Speed	Knowledge	Decision Making	Ergonomics	Safety
Identification & Handling	The collaborative system identifies the correct component from the feed area and moves the selected component to the work area.	Target Component		●	●
		Component Feeding		●	●
		Component Defects Detection		●		●
		Component pre-positioning		●	●		●	●
		Handling Prompting		●	●
		Balance Assistance	●	●			●	●
		Gripping Optimization	●	●
		Handling Instructions	●	●	●
		Tool Handover	●	●	●	●
		Data Logging			●	●
Alignment	The collaborative system aligns and places the component in its designated mounting location.	Target Component Positioning	●	●	●
		Subassembly Pre-Orientation		●	●		●	●
		Active Alignment Stabilization	●	●
		Precision Positioning	●	●			●
		Alignment Prompting		●	●
		Alignment Instructions	●	●	●
		Manual Alignment Assistance		●	●		●	●
Joining	The collaborative system physically joins the components.	Synchronized Component Joining	●	●	●
		Secure Holding	●	●			●
		Subassembly Pre-orientation		●	●		●	●
		Joining Prompting		●	●
		Joining Instructions	●	●	●
		Tool Handover	●	●	●	●
		Data Logging			●	●
Checking	The collaborative system performs a quality assessment to confirm that the assembly complies with specifications.	Collaborative system Error Mitigation		●	●
		Product Orientation for Inspection		●	●		●	●
		Inspection Instructions	●	●	●
		Checking Prompting		●	●
		Defect Tracking Assistance			●	●
		Tool Handover	●	●	●	●
Adjustment	The collaborative system rectifies defects in the assembled product to ensure that the final product meets specifications.	Product Orientation for Adjustment		●	●		●	●
		In-Process Validation	●	●		●
		Adjustment Instructions	●	●	●
		Real-Time Adjustment Guidance		●	●		●	●
		Adjustment Prompting		●	●
		Tool Handover	●	●	●	●

A case study applying the proposed methodology to skateboard assembly tasks in a collaborative assembly station is presented in this section. The skateboard assembly was selected as an example due to its relative simplicity, diverse range of components, and the requirement for various types of assembly task to be performed. Figure 3.A provides a graphical representation of the skateboard, outlining its complete assembled state. Figure 3.B breaks down the skateboard into its various components.

The case study considers a collaborative system equipped with a dual arm cobot holding two-finger grippers (see Fig. 4). A key feature of this station is a tracer laser projector, which offers an advanced way to guide the human operator by projecting graphical informative elements onto the workspace (e.g. which component to pick or where to place the component). Additionally, the assembly station is equipped with a vision system and a graphical user interface, accessible via a touch screen monitor.

The team responsible for the symbiotic task design comprised three individuals.

To deploy the proposed methodology in skateboard assembly, the main inputs are the B.O.M. (see Table 5) and a comprehensive ordered list of assembly tasks (see an extract in Table 6). The list of elementary tasks required to assemble the skateboard follows the taxonomy outlined in Section 5.2. This taxonomy categorizes the tasks into five distinct types: identification and handling, alignment, joining, checking, and adjustment. A total of 107 elementary tasks were delineated, providing a first roadmap from individual components to the completed skateboard assembly.

Table 5

Bill of Material of the assembled product in the case study. Note: 'LEVEL' indicates the hierarchy of each component within the product's structure. 'CODE' refers to the unique identifier assigned to each component. 'DESCRIPTION' provides an explanation or name of each component. 'QUANTITY' specifies the number of each component required in the product.
Level	Code	Description	Quantity
1	C	Base plate	2
1	C1	Kingpin	2
1	RB	Top cup	2
1	GB	Bushing	2
1	RS	Lower cup	2
1	GS	Cone bushing	2
1	GN	Pivot bushing	2
1	A	Hanger	2
1	D1	Kingpin nut	2
1	C2	Deck bolt	8
1	D2	Deck nut	8
1	R	Wheel	4
1	D3	Axle nut	4
1	T	Skateboard deck	1
2	ST	Sub-assembled truck	2
3	CP	Complete product	1

Table 6

List of elementary tasks for the assembly of the skateboard product. Note: Only an excerpt of the 107 elementary tasks required for assembly of the considered product is reported.
Order	Component code	task TYPE	Component description
1	B	Identification and handling	Base plate
2	C1	Identification and handling	Kingpin
3	C1	Alignment	Kingpin
4	RB	Identification and handling	Top cup
5	RB	Alignment	Top cup
…	…	…	…
36	T	Identification and handling	Skateboard deck
37	ST	Identification and handling	Sub-assembled truck
38	ST	Alignment	Sub-assembled truck
39	C2	Identification and handling	Deck bolt
40	C2	Alignment	Deck bolt
…	…	…	…
102	R	Alignment	Wheel
103	D3	Identification and handling	Axle nut
104	D3	Joining	Axle nut
105	D3	Checking	Axle nut
106	D3	Adjustment	Axle nut
107	CP	Identification and handling	Complete product

The following paragraphs will delve into how the tasks were categorized, allocated, and performed, showcasing the practical application of the method in an real-world setting.

Step 1: Task categorization

Tasks were categorized accordingly to two aspects: the elementary task type and the characteristic of grippability of the components, i.e., whether a component can be effectively grasped by the cobots’ grippers (Boothroyd, 1994; Malik and Bilberg, 2019). Due to the limitations of the existing two-finger grippers in handling certain types of components, this feature was considered critical. In this specific application case, no other properties of components were considered relevant.

To ensure a comprehensive evaluation of grippability, components in the skateboard assembly process were analyzed against five key factors: size and thickness, weight, shape, stability, and sensitivity (Boothroyd, 1994; Malik and Bilberg, 2019). Components failing to meet any one of these criteria were classified as 'ungrippable,' by the cobot.

The assessments for categorization were conducted by the development team. They based their evaluations on the specific features of each component and the capabilities of the collaborative system (Malik and Bilberg, 2019).

Table 7 illustrates this classification, correlating the physical attributes of the skateboard components with their grippability status.

Table 7

Component classification based on grippability criteria. Note: 'CODE' refers to the unique identifier for each component. 'DESCRIPTION' provides an explanation or name of each component. "COMPONENT CATEGORY' indicates the classification of the component into two categories: grippable and ungrippable. "YES" and "NO" refers to whether or not the Grippability criteria as defined by Malik and Bilberg (2019) are met.
CODE	DESCRIPTION	GRIPPABILITY CRITERIA					COMPONENT CATEGORY
CODE	DESCRIPTION	Size and thickness	Weight	Shape	Stability	Sensitivity	COMPONENT CATEGORY
B	Base plate	YES	YES	YES	YES	YES	Grippable
C1	Kingpin	YES	YES	NO	YES	YES	Ungrippable
RB	Top cup	NO	YES	NO	YES	YES	Ungrippable
GB	Bushing	NO	YES	NO	NO	YES	Ungrippable
RS	Lower cup	NO	YES	NO	YES	YES	Ungrippable
GS	Cone bushing	NO	YES	NO	NO	YES	Ungrippable
GN	Pivot bushing	NO	YES	NO	NO	NO	Ungrippable
A	Hanger	YES	YES	NO	YES	YES	Ungrippable
D1	Kingpin nut	NO	YES	NO	YES	YES	Ungrippable
C2	Deck bolt	NO	YES	NO	YES	YES	Ungrippable
D2	Deck nut	NO	YES	NO	YES	YES	Ungrippable
R	Wheel	YES	YES	YES	YES	YES	Grippable
D3	Axle nut	NO	YES	NO	YES	YES	Ungrippable
T	Skateboard deck	YES	YES	YES	YES	YES	Grippable
ST	Sub-assembled truck	YES	YES	YES	YES	YES	Grippable
CP	Complete product	YES	YES	YES	YES	YES	Grippable

The combination of grippability properties with the elementary task types - referenced in Table 6 - resulted in the identification of seven task categories in the skateboard assembly process: (i) identification and handling tasks for grippable components; (ii) identification and handling tasks for ungrippable components; (iii) alignment tasks for grippable components; (iv) alignment tasks for ungrippable components; (v) joining tasks for ungrippable components; (vi) checking tasks for ungrippable components and (vii) adjustment tasks for ungrippable components. Some elementary types of tasks are associated with both grippable and ungrippable components, reflecting their relevance to both types of parts. Other types of tasks are only associated with the ungrippable category.

Step 2: Symbiotic task allocation

The second step involved assigning the roles of leader or supporter to either the human operator or the collaborative system for each task category.

Adopting the task allocation approach described in Section 5.3, the development team evaluated each task category against the six symbiotic dimensions: effort, speed, knowledge, decision making, ergonomics and safety, as detailed in Table 2. This comprehensive analysis aimed to identify the most appropriate leader or supporter for each task category, in order to maximize symbiotic interactions and increase the efficiency of HRC.

Evaluations and the outcomes of this task allocation process are presented in Table 8. Notably, it was determined that the 'Identification and handling of Grippable' components category was optimally leaded by the collaborative system. Conversely, for all other task categories, the human operator emerged as the preferable leader. The supporter role, in each case, complements the leader by assisting in task execution, ensuring that the collaboration is smooth, and the symbiotic potential is exploited.

Table 8

Symbiotic Task allocation application. Note: 'TASK CATEGORY' lists the task categories identified in the task categorization phase. 'SYMBIOTIC COMPONENTS’ refers to the criteria used for evaluating the potential for maximizing symbiotic exchanges, as outlined in Section 2. The columns 'LEADER' and 'SUPPORTER' indicate the assignments of agents to these two roles for each task category using the allocation approach outlined in Section 5.3. Regarding the evaluation of 'SYMBIOTIC DIMENSIONS', the cells contain responses to the questions presented in Table 2 ad detailed in Section 5.3 (N.A. = Not Applicable).
TASK CATEGORY	SYMBIOTIC COMPONENTS						LEADER	SUPPORTER
TASK CATEGORY	Effort	Speed	Knowledge	Decision making	Ergonomics	Safety	LEADER	SUPPORTER
Identification and handling grippable	Collab. System	Human	Collab. System	N.A.	Collab. System	Collab. System	Collaborative system	Human
Identification and handling ungrippable	Human	Human	Collab. System	N.A.	Equal	Equal	Human	Collaborative system
Alignment grippable	Human	Human	Human	Human	N.A.	N.A.	Human	Collaborative system
Alignment ungrippable	Human	Human	Human	Human	N.A.	N.A.	Human	Collaborative system
Joining ungrippable	Human	Human	Human	N.A.	N.A.	N.A.	Human	Collaborative system
Checking ungrippable	Human	Human	Human	Human	N.A.	N.A.	Human	Collaborative system
Adjustment ungrippable	Human	Human	Human	Human	N.A.	N.A.	Human	Collaborative system

Step 3: Task protocol generation

This phase focuses on outlining how the tasks should be performed. Initially, each task category was associated with specific activities that need to be performed by the leader and the supporter. Tables 3 and 4, which detail the actions required for each task category, have been used to facilitate this mapping. When selecting supporting activities, several factors were taken into account, including the capabilities of the collaborative system, the ergonomic requirements of the human operators and the efficiency of the assembly process, with the aim of selecting actions capable of enhancing the leader's skills and promoting symbiotic exchange.

Basic activities listed in Tables 3 and 4 were specified to match the technological capabilities of the collaborative system under analysis. This step was crucial for effectively translating the conceptual framework into practical, technology-specific applications.

Table 9 reports a breakdown of the activities undertaken by both the leader and supporter for each task category, offering a structured perspective on how each task category should be approached in the collaborative assembly process.

The complete task protocol was efficiently developed by referring to Table 9, which facilitated the alignment of each task category in the skateboard assembly process with the corresponding activities of the human operator and collaborative system. An excerpt of the complete task protocol is presented in Table 10. This table provides a snapshot of the dynamic interplay between the human and collaborative system across various stages of the assembly process and highlights the specific actions undertaken by the human and collaborative system for each task.

At the end of the design process, it is noteworthy to consider how the complexity of the problem was managed. What began as an extensive consideration of 107 distinct assembly tasks was condensed to just 7 consistent and homogenous task categories. Design efforts were focused on this limited number of task categories. Finally, the solutions identified from the design of the 7 task categories were systematically applied to the original 107 tasks. This approach ensured efficiency and consistency throughout the assembly process.

Table 9

Results of the task protocol development for the skateboard case study.
TASK CATEGORY	ALLOCATION LEADER	ALLOCATION SUPPORTER	LEADER ACTIVITY	SUPPORTIVE ACTIVITY
Identification and handling Grippable	Collaborative system	Human	The collaborative system identifies the component in the feeding area with the vision system and moves the component with the robotic arm 1	Handling prompting: the human operator tells the collaborative system when to proceed with component handling via a trigger on the interface screen Component pre-positioning: if the component is not in the optimal position, the human operator corrects the component's position.
Identification and handling Ungrippable	Human	Collaborative system	The operator identifies and handles the component	Target component: the vision system identifies the position of the component and through the projector laser tracer, the collaborative system highlights the component to be handled by the human operator
Alignment grippable	Human	Collaborative system	The operator correctly positions the component	Subassembly pre-orientation: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment
Alignment ungrippable	Human	Collaborative system	The operator correctly position the component	Secure holding: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment
Joining ungrippable	Human	Collaborative system	The operator performs the joining activity on the component	Tool handover: the collaborative system with robotic arm 2 brings the tool needed to perform the operation closer to the operator Secure holding: the collaborative system with robotic arm 1 holds the part in a fixed and comfortable position to perform the assembly activities Joining instructions: through the graphic interface the collaborative system shows the instructions on correct joining
CHECKING UNGRIPPABLE	Human	Collaborative system	The operator checks the correct assembly of the component	Secure holding: the collaborative system with robotic arm 1 holds the workpiece in a fixed and comfortable position to perform assembly tasks Visual field optimisation: the collaborative system illuminates through the projector the areas to be controlled Adaptive ergonomic product positioning: the collaborative system with robotic arm 1 moves the component so as to bring the areas to be checked closer to the operator Checking instructions: through the graphic interface the collaborative system shows the instructions on correct checking
ADJUSTMENT UNGRIPPABLE	Human	Collaborative system	The operator adjusts the assembly of the component	Tool handover: the collaborative system with robotic arm 2 brings the tool required to perform the operation closer to the operator Secure holding: the collaborative system with robotic arm 1 holds the workpiece in a fixed and comfortable position to perform the adjustment tasks Adjustment instructions: through the graphic interface the collaborative system shows the instructions on correct adjustment

Table 10

Assembly task protocol. Note: Only an excerpt of the 107 elementary tasks required for assembly of the considered product has been reported.
ORDER	COMPONENTS CODE	TASK CATEGORY	LEADER	SUPPORTER	LEADER ACTIVITY	SUPPORTIVE ACTIVITY
1	B	Identification and handling Grippable	COLLABORATIVE SYSTEM	HUMAN	The collaborative system identifies and moves the component B with the robotic arm 1	Handling prompting: the human operator indicates to the collaborative system when to proceed with the handling of the component B via a trigger on the interface screen + Component pre-positioning: if the component is not in the optimal position, the human operator corrects the component's position
2	C1	Identification and handling Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator identifies and handles the component C1	Target component: the vision system identifies the position of the component C1 and through the projector laser tracer, the collaborative system highlights the component C1 to be handled by the human operator
3	C1	Alignment Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator correctly positions the component C1	Secure holding: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities + Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment of component C1
4	RB	Identification and handling Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator identifies and handles the component RB	Target component: the vision system identifies the position of the component RB and through the projector laser tracer, the collaborative system highlights the component RB to be handled by the human operator
5	RB	Alignment Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator correctly positions the component RB	Secure holding: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities + Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment
…	…	…	…	…	…	…
36	T	Identification and handling Grippable	COLLABORATIVE SYSTEM	HUMAN	The collaborative system identifies and moves the component T	Handling prompting: the human operator indicates to the collaborative system when to proceed with the handling of the component T via a trigger on the interface screen + Component pre-positioning: if the component is not in the optimal position, the human operator corrects the component's position.
37	ST	Identification and handling Grippable	COLLABORATIVE SYSTEM	HUMAN	The collaborative system identifies and moves the Sub-assembled truck	Handling prompting: the human operator indicates to the collaborative system when to proceed with the handling of the component ST via a trigger on the interface screen + Component pre-positioning: if the component is not in the optimal position, the human operator corrects the component's position.
38	ST	Alignment Grippable	HUMAN	COLLABORATIVE SYSTEM	The operator correctly positions the Sub-assembled truck	Subassembly pre-orientation: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities + Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment
39	C2	Identification and handling Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator identifies and handles the component C2	Target component: the vision system identifies the position of the component C2 and through the projector laser tracer, the collaborative system highlights the component C2 to be handled by the human operator
40	C2	Alignment Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator correctly positions the component C2	Secure holding: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities + Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment
…	…	…	…	…	…	…
102	R	Alignment Grippable	HUMAN	COLLABORATIVE SYSTEM	The operator correctly positions the component R	Subassembly pre-orientation: the collaborative system with robotic arm 1 conveniently holds the part in a position to perform assembly activities + Alignment instructions: through the graphic interface the collaborative system shows instructions on correct alignment
103	D3	Identification and handling Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator identifies and handles the component D3	Target component: the vision system identifies the position of the component D3 and through the projector laser tracer, the collaborative system highlights the component D3 to be handled by the human operator
104	D3	Joining Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator performs the joining activity on the component D3	Tool handover: the collaborative system with robotic arm 2 brings the tool needed to perform the operation closer to the operator + Secure holding: the collaborative system with robotic arm 1 holds the part in a fixed and comfortable position to perform the assembly activities + Joining instructions: through the graphic interface the collaborative system shows the instructions on correct joining
105	D3	Checking Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator checks the correct assembly of the component D3	Secure holding: the collaborative system with robotic arm 2 holds the part in a fixed and comfortable position to perform assembly tasks + Visual field optimisation: the collaborative system illuminates through the projector the areas to be controlled + Adaptive ergonomic product positioning: the collaborative system with robotic arm 1 moves the part so as to bring the areas to be checked closer to the operator. + Checking instructions: through the graphic interface the collaborative system shows the instructions on correct checking
106	D3	Adjustment Ungrippable	HUMAN	COLLABORATIVE SYSTEM	The operator adjusts the assembly of the component D3	Tool handover: the collaborative system with robotic arm 2 brings the tool required to perform the operation closer to the operator. + Secure holding: the collaborative system with robotic arm 1 holds the workpiece in a fixed and comfortable position to perform the adjustment tasks + Adjustment instructions: through the graphic interface the collaborative system shows the instructions on correct adjustment
107	CP	Identification and handling Grippable	COLLABORATIVE SYSTEM	HUMAN	The collaborative system identifies and moves complete product CP with the robotic arm 1	Handling prompting: the human operator indicates to the collaborative system when to proceed with the handling of the final product CP via a trigger on the interface screen + Component pre-positioning: if the component is not in the optimal position, the human operator corrects the component's position.

This article proposed a novel methodology for enhancing collaborative assembly processes through the design of symbiotic task. The ultimate goal of this method is to develop collaborative processes that maximize symbiotic exchanges between humans and cobots.

The proposed methodology starts with a comprehensive analysis of the activities involved in an assembly process, decomposing complex processes into consistent task categories, which simplifies the problem domain and addresses the core questions of what actions are necessary to accomplish each task. Following this categorization, the process involves the allocation of leader and supporter roles for each elementary task, with the aim of aligning these roles to enhance the symbiotic exchanges between collaborators. The third and final step involves the generation of the specific task protocol. This stage is where the behavior of the collaborative agents is outlined, providing a clear and structured approach to how the tasks should be executed. For the design of the task protocol, a list of supportive activities impacting on specific components of Symbiotic Human-Robot Collaboration (SHRC) is proposed to guide the process.

This methodology highlights the importance of developing a deeper synergy between human operators and collaborative systems in assembly tasks. It is through this integration that the potential of collaborative robotics can be realized and more widely adopted. The proposed approach suggests that improving collaboration involves more than just task allocation; it requires creating an environment where humans and collaborative systems can mutually support and enhance each other's work. By fostering a scenario where each agent not only performs tasks but also assists the other, the methodology aims to create a more cohesive and harmonious collaborative workspace, thereby improving overall process outcomes. Moreover, by improving the symbiotic relationship between human operators and robots, the methodology also focuses on elevating the well-being and ergonomic comfort of the human workforce, aligning with the principles of Human-Centric Manufacturing.

However, it is important to acknowledge the limitations inherent in the proposed methodology. Firstly, a limitation of this study is the specificity of the identified supportive activities, which are primarily oriented for assembly processes. While the underlying principles of the methodology are adaptable and can potentially be applied to a wide range of manufacturing settings, this adaptation requires the identification of new supportive activities. Secondly, while the methodology provides a guide for SHRC design, it also allows for considerable freedom in decision-making. This flexibility can lead to different implementations based on the user's experience. Thirdly, decomposing complex processes into elementary tasks can result in the creation of highly detailed task protocols. These limitations highlight areas for potential refinement and underscore the importance of user expertise and judgment in applying the methodology effectively.

Looking to the future, a promising direction for advancement is the incorporation of emerging AI and machine learning technologies to further enhance the decision-making process in the design of symbiotic assembly task. Furthermore, investigating the potential for expanding the applicability of this approach to more complex assembly processes and different industrial contexts would be valuable.

Looking to the future, a promising direction for advancement lies in the incorporation of AI and machine learning technologies to further enhance the decision-making process in the design of symbiotic assembly tasks. This may also involve developing tools to effectively identify the most appropriate support activities among different options. Furthermore, investigating the potential for expanding the applicability of this approach to more complex assembly processes and different industrial contexts would be valuable. Such exploration could lead to broader adaptability and effectiveness of this methodology across a range of manufacturing scenarios.

Ethical Approval

The authors respect the Ethical Guidelines of the Journal.

Consent to Participate

Not applicable

Consent to Publish

Not applicable

Authors Contributions

The authors have provided an equal contribution to the drafting of the paper.

Funding

This work has been partially supported by the European Union – NextGenerationEU.

Competing Interests

The authors do not have conflict of interest

Availability of data and materials

Not applicable

Antonelli, D., Bruno, G., 2019. Dynamic distribution of assembly tasks in a collaborative workcell of humans and robots, FME Transactions, 47(4), 723–730.
Bänziger, T., Kunz, A., Wegener, K., 2020. Optimizing human–robot task allocation using a simulation tool based on standardized work descriptions, Journal of Intelligent Manufacturing, 31(7), 1635–1648.
Barravecchia, F., Bartolomei, M., Mastrogiacomo, L., Franceschini, F., 2023. Advancing Human-Robot Collaboration: proposal of a methodology for the design of Symbiotic Assembly Workstations, Procedia Computer Science: Proceedings of International Conference on Industry 4.0 and Smart Manufacturing (ISM), Lisbon, November 2023, Article in Press.
Barravecchia, F., Bartolomei, M., Mastrogiacomo, L., Franceschini, F., 2023a. Redefining Human–Robot Symbiosis: a bio-inspired approach to collaborative assembly, International Journal of Advanced Manufacturing Technology, 128(5–6), 2043 – 2058.
Barravecchia, F., Mastrogiacomo, L., Franceschini, F., 2023b. A general cost model to assess the implementation of collaborative robots in assembly processes, International Journal of Advanced Manufacturing Technology, 125(11–12), 5247 – 5266.
Boothroyd, G., 1994. Product design for manufacture and assembly, Computer-Aided Design, 26(7), 505–520.
Boothroyd, G., Dewhurst, P., Knight, W.A., 2010. Product Design for Manufacture and Assembly: Third Edition, Taylor & Francis Group, Broken Sound Parkway NW (US).
Borboni, A., Reddy, K.V.V., Elamvazuthi, I., AL-Quraishi, M.S., Natarajan, E., Azhar Ali, S.S., 2023. The Expanding Role of Artificial Intelligence in Collaborative Robots for Industrial Applications: A Systematic Review of Recent Works, Machines, 11(1), 1 – 28.
Cesta, A., Orlandini, A., Umbrico, A., 2018. Fostering Robust Human-Robot Collaboration through AI Task Planning, Procedia CIRP, 72, 1045–1050.
Cinelli, M., Kadziński, M., Gonzalez, M., Słowiński, R., 2020. How to support the application of multiple criteria decision analysis? Let us start with a comprehensive taxonomy, Omega (United Kingdom), 96, 102261.
De Montis, A., De Toro, P., Droste-Franke, B., Omann, I., Stagl, S., 2004. Assessing the quality of different MCDA methods, in: Alternatives for Environmental Valuation. Taylor and Francis Inc., pp. 99–133. ISBN: 9780203412879.
Dmytriyev, Y., Insero, F., Carnevale, M., Giberti, H., 2022. Brain–Computer Interface and Hand-Guiding Control in a Human–Robot Collaborative Assembly Task, Machines, 10(654), 1 – 16.
Faccio, M., Cohen, Y., 2023. Intelligent cobot systems: human-cobot collaboration in manufacturing, Journal of Intelligent Manufacturing, Article in Press.
Faccio, M., Granata, I., Menini, A., Milanese, M., Rossato, C., Bottin, M., Minto, R., Pluchino, P., Gamberini, L., Boschetti, G., Rosati, G., 2023. Human factors in cobot era: a review of modern production systems features, Journal of Intelligent Manufacturing, 34(1), 85 – 106.
Fiore, M., Clodic, A., Alami, R., 2016. On planning and task achievement modalities for human-robot collaboration, Springer Tracts in Advanced Robotics, 109, 293–306.
Garcia, P.P., Santos, T.G., Machado, M.A., Mendes, N., 2023. Deep Learning Framework for Controlling Work Sequence in Collaborative Human–Robot Assembly Processes, Sensors, 23(553), 1 – 18.
Gervasi, R., Barravecchia, F., Mastrogiacomo, L., Franceschini, F., 2022a. Applications of affective computing in human-robot interaction: State-of-art and challenges for manufacturing, Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, 237(6–7), 815–832.
Gervasi, R., Mastrogiacomo, L., Franceschini, F., 2020. A conceptual framework to evaluate human-robot collaboration, International Journal of Advanced Manufacturing Technology, 108(3), 841–865.
Gervasi, R., Mastrogiacomo, L., Maisano, D.A., Antonelli, D., Franceschini, F., 2022b. A structured methodology to support human–robot collaboration configuration choice, Production Engineering, 16(4), 435 – 451.
Huang, Y., Leung, P.Y.V., Garrett, C., Gramazio, F., Kohler, M., Mueller, C., 2021. The new analog: A protocol for linking design and construction intent with algorithmic planning for robotic assembly of complex structures, Proceedings - SCF 2021: ACM Symposium on Computational Fabrication.
Huang, Y.M., Shiau, C.S., 2009. An optimal tolerance allocation model for assemblies with consideration of manufacturing cost, quality loss and reliability index, Assembly Automation, 29(3), 220–229.
Inkulu, A.K., Bahubalendruni, M.V.A.R., Dara, A., K., S., 2022. Challenges and opportunities in human robot collaboration context of Industry 4.0 - a state of the art review, Industrial Robot: the international journal of robotics research and application, 49(2), 226–239.
Ishizaka, A., Nemery, P., 2013. Multi-Criteria Decision Analysis: Methods and Software, Multi-Criteria Decision Analysis: Methods and Software. John Wiley & Sons Ltd, Chichester (UK).
ISO 13732-1:2006, Ergonomics of the thermal environment—Methods for the assessment of human responses to contact with surfaces—Part 1: Hot surfaces. Geneva, 2008.
Kiyokawa, T., Shirakura, N., Wang, Z., Yamanobe, N., Ramirez-Alpizar, I.G., Wan, W., Harada, K., 2023. Difficulty and complexity definitions for assembly task allocation and assignment in human–robot collaborations: A review, Robotics and Computer-Integrated Manufacturing, 84, 102598.
Lee, M.L., Behdad, S., Liang, X., Zheng, M., 2022. Task allocation and planning for product disassembly with human–robot collaboration, Robotics and Computer-Integrated Manufacturing, 76, 102306.
Li, S., Wang, R., Zheng, P., Wang, L., 2021. Towards proactive human–robot collaboration: A foreseeable cognitive manufacturing paradigm, Journal of Manufacturing Systems, 60, 547–552.
Liau, Y.Y., Ryu, K., 2022. Genetic algorithm-based task allocation in multiple modes of human–robot collaboration systems with two cobots, International Journal of Advanced Manufacturing Technology, 119(11–12), 7291–7309.
Lihui Wang, S.K., Feng, H.-Y., 2011. A function block based approach for increasing adaptability of assembly planning and control, International Journal of Production Research, 49(16), 4903–4924.
Liu, H., Qu, D., Xu, F., Zou, F., Song, J., Jia, K., 2019. A Human-Robot Collaboration Framework Based on Human Motion Prediction and Task Model in Virtual Environment, 9th IEEE International Conference on Cyber Technology in Automation, Control and Intelligent Systems, CYBER 2019, 1044–1049.
Lu, Y., Zheng, H., Chand, S., Xia, W., Liu, Z., Xu, X., Wang, L., Qin, Z., Bao, J., 2022. Outlook on human-centric manufacturing towards Industry 5.0, Journal of Manufacturing Systems, 62, 612–627.
Malik, A.A., Bilberg, A., 2019. Complexity-based task allocation in human-robot collaborative assembly, Industrial Robot, 46(4), 471–480.
Mhenni, F., Vitolo, F., Rega, A., Plateaux, R., Hehenberger, P., Patalano, S., Choley, J.Y., 2022. Heterogeneous Models Integration for Safety Critical Mechatronic Systems and Related Digital Twin Definition: Application to a Collaborative Workplace for Aircraft Assembly, Applied Sciences (Switzerland), 12(6).
Michalos, G., Spiliotopoulos, J., Makris, S., Chryssolouris, G., 2018. A method for planning human robot shared tasks, CIRP Journal of Manufacturing Science and Technology, 22, 76–90.
Mulesa, O., Horvat, P., Radivilova, T., Sabadosh, V., Baranovskyi, O., Duran, S., 2023. Design of mechanisms for ensuring the execution of tasks in project planning, Eastern-European Journal of Enterprise Technologies, 2(4–122), 16–22.
Müller, R., Vette, M., Mailahn, O., 2016. Process-oriented Task Assignment for Assembly Processes with Human-robot Interaction, Procedia CIRP, 44, 210–215.
Navas-Reascos, G.E., Romero, D., Rodriguez, C.A., Guedea, F., Stahre, J., 2022. Wire Harness Assembly Process Supported by a Collaborative Robot: A Case Study Focus on Ergonomics, Robotics, 11(131), 1–18.
Petzoldt, C., Harms, M., Freitag, M., 2023. Review of task allocation for human-robot collaboration in assembly, International Journal of Computer Integrated Manufacturing, 36(11), 1675–1715.
Pupa, A., Secchi, C., 2021. A Safety-Aware Architecture for Task Scheduling and Execution for Human-Robot Collaboration, IEEE International Conference on Intelligent Robots and Systems, 1895–1902.
Radkowski, R., Herrema, J., Oliver, J., 2015. Augmented Reality-Based Manual Assembly Support With Visual Features for Different Degrees of Difficulty, International Journal of Human-Computer Interaction, 31(5), 337–349.
Ranz, F., Hummel, V., Sihn, W., 2017. Capability-based Task Allocation in Human-robot Collaboration, Procedia Manufacturing, 9, 182–189.
Reid, R.D., Sanders, N.R., 2020. Operations Management: An Integrated Approach, 7th ed. John Wiley & Sons Ltd, United States of America.
Simões, A.C., Pinto, A., Santos, J., Pinheiro, S., Romero, D., 2022. Designing human-robot collaboration (HRC) workspaces in industrial settings: A systematic literature review, Journal of Manufacturing Systems, 62, 28–43.
Tamminen, P., Ukkonen, L., Sydänheimo, L., 2016. Correlation of component human body model and charged device model qualification levels with electrical failures in electronics assembly, Journal of Electrostatics, 79, 38–44.
Tsarouchi, P., Makris, S., Chryssolouris, G., 2016. Human–robot interaction review and challenges on task planning and programming, International Journal of Computer Integrated Manufacturing, 29(8), 916–931.
Tsarouchi, P., Michalos, G., Makris, S., Athanasatos, T., Dimoulas, K., Chryssolouris, G., 2017. On a human–robot workplace design and task allocation system, International Journal of Computer Integrated Manufacturing, 30(12), 1272–1279.
Verna, E., Puttero, S., Genta, G., Galetto, M., 2023. Exploring the Effects of Perceived Complexity Criteria on Performance Measures of Human-Robot Collaborative Assembly, Journal of Manufacturing Science and Engineering, 145(10), 101104.
Wang, L., Gao, R., Váncza, J., Krüger, J., Wang, X. V., Makris, S., Chryssolouris, G., 2019. Symbiotic human-robot collaborative assembly, CIRP Annals, 68(2), 701–726.
Wang, X.V., Kemény, Z., Váncza, J., Wang, L., 2017. Human–robot collaborative assembly in cyber-physical production: Classification framework and implementation, CIRP Annals - Manufacturing Technology, 66(1), 5 – 8.

APPENDIXA.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Integrating Bio Principles in Assembly Task Design to Enhance Symbiotic Human-Robot Collaboration

Status:

Version 1

Abstract

Figures

1 INTRODUCTION

2 HUMAN-ROBOT SYMBIOSIS

3 LITERATURE REVIEW ON COLLABORATIVE TASK ALLOCATION

4 RESEARCH OVERVIEW ON COLLABORATIVE TASK DESIGN

5 DESIGN OF SYMBIOTIC TASK

5.1 Input of the method

Step 1: Task categorization

Step 2: Symbiotic task allocation

Step 3: Task protocol generation

6 CASE STUDY

Step 1: Task categorization

Step 2: Symbiotic task allocation

Step 3: Task protocol generation

DISCUSSIONS AND CONCLUSIONS

Declarations

References

Supplementary Files

Status:

Version 1