Bridging the collectives: A review of collective human

Abstract

Advancements in multi-agent, autonomous, and intelligent robotic systems over the past decades point toward new design and fabrication possibilities. Exploring how humans and robots can create and construct collectively is essential in leveraging robotic technology in the building sector. However, only by making existing knowledge from relevant technological disciplines accessible to designers can we fully exploit current construction methods and further develop them to address the challenges in architecture. To do this, we present a review paper that bridges the gap between Collective Robotic Construction (CRC) and Human–Robot Interaction (HRI) and defines a new research domain in Collective Human–Robot Construction (CHRC) in the architectural design and fabrication context.

Keywords

collective human–robot construction collective robotic construction human–robot interaction robotic fabrication human–robot teams

Introduction

Developments in robotic fabrication have inspired new methods of architectural construction.¹ While most existing fabrication processes rely on the use of singular machines, recent developments in the field have focused increasingly on methods of multi-robotic fabrication and the advantages these bring to design and construction.² This shift from one to multiple machines requires not only rethinking the robotic process itself but also developing more dynamic and flexible methods for robotic control.

Processes involving multiple machines may be aimed at increasing a machine’s reach, speed, and efficiency. In addition to these goals, several projects investigate how multiple machines may support each other or take on individual tasks that lead to new construction processes and outcomes.^3–6 As a result, machine interaction and control in multi-robotic processes have developed into a central area of contemporary research in robotic fabrication.

Even though crucial to the success of robotic fabrication processes, the role of humans in these processes has often not been the primary concern. Humans were viewed as users or designers pre-defining robotic tasks. However, even in the simplest robotic processes, humans are directly involved. They often perform as more than just the executors of tasks by monitoring the robotic process, making necessary adjustments, and improving the construction procedures on the fly. We can view robotic construction processes from two perspectives: (1) full automation aimed at replacing humans and (2) synergistic processes between humans and machines that expand available design and construction possibilities. While automation might lead to faster and more efficient construction processes, taking advantage of human agents in the fabrication process is a much more valuable and sustainable goal. Instead of reducing human workers’ presence on construction sites, it is worthwhile to explore how humans can be supported by machines and vice-versa. Therefore, the interaction between humans and robots, along with the control of multiple machines, becomes a crucial research area for the further development of robotic construction.

Human–machine interaction and machine–machine collaboration are not new research areas. Although they have only recently found their way into architecture and design, they are building upon a substantial body of knowledge from other disciplines such as robotics, mechanical engineering, and computer science. Architectural research has the potential to link multiple such disciplines through new approaches and applications. Therefore, it is essential for architects and designers to have access to the knowledge outside our field, in this case specifically that of robotics. By making existing methods and developments accessible to designers, we can identify new areas of research at the intersections of robotics-related disciplines and architecture, which would not be accessible by addressing each topic individually.

Method

We have identified the fields of Collective Robotic Construction (CRC) and Human–Robot Interaction (HRI) as central to our investigation since they cover researches from both robotic and human perspectives in the construction process. CRC focuses on multi-robotic systems collectively modifying a shared environment^2,7 while HRI is dedicated to defining and crafting the relationship between humans and robotic systems.^8–11 Our review aims to bridge these two areas and identify a new research domain in Collective Human–Robot Construction (CHRC).

This paper is structured as follows. Firstly, we give a critical survey of existing works in CRC and HRI. Then, we set up two principal research axes (ranging from autonomy to collaboration, from design to fabrication) to situate and connect major topics in CRC and HRI. Finally, we locate research gaps between CRC and HRI and highlight near-future opportunities in the newly defined field, CHRC. As a result, this paper is intended to be a review paper that goes beyond the depiction of existing research but critically highlights research gaps and resulting new opportunities.

Review

Collective Robotic Construction (CRC) concerns “embodied, autonomous, multirobot systems that modify a shared environment according to high-level user-specified goals,” as defined by Petersen et al.² in 2019. This is an emerging field (Figure 1) that integrates research on multi-agent systems, which explores the distributed operation of autonomous agents in a network, into construction robotics, which focuses on the automation of construction processes through the use of robots.¹² According to Petersen et al., CRC can be a critical tool to enhance the scalability and adaptability of robotic fabrication in the architectural field by integrating design, construction, mechanism, and control.² However, CRC examples, such as termite-inspired robot construction,¹³ tensile web construction,¹⁴ and micro-robot truss construction,¹⁵ are mainly conducted in controlled lab environments at the moment, with humans often acting as operators for robots. Thus, there is research space to further explore alternative ways in which humans can interact with such swarm robotic systems to create new fabrication processes or make existing robotic construction methods more transferable to dynamic construction sites.

Figure 1.

The emerging field of CRC is at the interaction of many existing fields. Common research themes are highlighted in the corners. Illustration by Petersen et al.² in 2019.

Human–Robot Interaction (HRI) focuses on “understanding, designing, and evaluating robotic systems for use by or with humans,” as stated in Goodrich’s survey in 2008.⁹ The literature on HRI has expanded rapidly over the past 15 years.¹⁶ It consists of broad and diverse topics, which can be loosely organized into five major categories: navigation, perception, management, manipulation, and social.¹⁰ The rich literature in HRI covers frameworks and methods for creating collaborative, interactive, and efficient working relationships between humans and machines.

As both CRC and HRI are young research domains that have become exponentially active within the past two decades, we expect to find previously un-discovered bridges between the two literature spaces. Besides, because humans and construction processes are inseparable in the building industry (both in terms of design and execution), exploring the topic between CRC and HRI can help better leverage robotic technologies in the overall building sector.

In the following subsections, we blend the notion of collectiveness into HRI and human factors into CRC. Collective agents open up new ways of collaborating and decision-making in design procedures (e.g., less centralized control, stigmergic mechanisms, or feedback-based processes) and material manipulation. Meanwhile, better integration of human operators into the robotic construction process accelerates technology transfer of robotics into real-world practice, expands fabrication methods by combining both parties’ strengths, and ties digital fabrication back to human scale and values. In addition, some shared inspirations drawn from biology and game theory for CRC and HRI are also discussed in this section. The intentional cross-referencing interweaves literature reviews between the two domains and aims to reveal potential gaps and opportunities.

Collectiveness

The term collectiveness implies a networked system linking individual agents into a whole. A collective or multi-agent system contains more than one agent to form a higher-level entity through interactions among the agents. Ferber defined the terms “agent” and “multi-agent system” as follows in 1999:

“An agent can be a physical or virtual entity that can act, perceive its environment (in a partial way) and communicate with others, is autonomous and has skills to achieve its goals and tendencies. It is in a multi-agent system (MAS) that contains an environment, objects and agents (the agents being the only ones to act), relations between all the entities, a set of operations that can be performed by the entities and the changes of the universe in time and due to these actions.”¹⁷

One major advantage of a collective system is that it can form large-scale complex entities without requiring each agent to perform complicated computations or actions. When it comes to physical power, a group of swarm robots can manipulate material elements that are much larger in scale than any individual robot can individually manage. Examples of such can be seen in emergent structures¹⁸ and distributed timber constructions.¹⁹ Computational-wise, collective intelligence can achieve significant complexity without requiring heavy information processing from each agent. Wisdom of the crowd is generated through a diverse and decentralized collective that supports individual independence and provides appropriate aggregation mechanisms.²⁰ Another implication of collectiveness particularly in architectural design is group creativity, which involves multiple agents coming up with new ideas and creations, often in an improvisational manner. Group creativity, or improvisational creativity, differs from the traditional product or compositional creativity due to its non-linear workflow and equal emphasis on process and final result.^21–23 Existing examples demonstrating the possibility of heterogeneous human–robot teams improvising together can be seen in music and theatrical performances (e.g., multi-robot–human jazz jam session^24,25 and theater improvisation²⁶). However, more experimental projects in human–robot improvisational construction are needed to assess the potential of robotics technology’s influence on group creativity in architectural design and fabrication.

Research interests in collective or multi-agent systems emerged way earlier than the prevalence of robotics, dating back to the 18th century.^27–29 Thus, many notions in the collective systems are not robot-specific (e.g., the Object Management Group,³⁰ Foundation for Intelligent Physical Agents,³¹ Knowledge-able Agent-oriented System,³² and General Magic’s Models^33,34). However, researchers should still be aware of such this area of research and active transfer general principles into heterogeneous human–robot teams in design–fabrication context.

The construction industry consists of highly collective human social structures, including but not limited to roles such as architects, engineers, construction workers, electricians, landscape designers, and interior designers. Because relationships amongst these professions can be influenced by technological, social, and economic factors, introducing a new tool such as the robot can often inflict relationship changes within construction-related occupations. For example, the recent prevalence of Building Information Modeling (BIM³⁵) has improved the workflow across design, engineering, construction, and maintenance. Similarly, the notions such as Robot-Oriented Design^36,37 were introduced to adjust the value chain “corresponding with information flows between enterprise, product, machinery, robots, costumers and all complementary sub-processes.³⁸” This helps the industry to better address social, environmental, and financial dimensions.

Human factors

Robotics and automation were initially invented to assist humans in completing tasks; hence, the human factor is an important subject to consider in order to create corresponding robotic systems and human–machine relationships.

The human factor, in relation to robots, can be interpreted in many ways. For example, humans can be trained (e.g., operators) or untrained (e.g., consumers and passerby), supportive or disruptive, individual or collective. Regardless of the specific types, humans generally make mistakes, perform less consistently, and are biased.⁴⁰ Thus, researchers can always draw references from psychology and sociology in building a fundamental understanding of human behaviors before designing a befitting robotic system. For the purpose of this review, we limit the role of humans to trained users involved in a construction project, such as designers, engineers, and construction workers.

HRI covers topics regarding how humans interact with robots. The three major components of HRI are defined as: Robot, Human, and System.¹⁰ In 2004, Yanco updated a list of taxonomy for HRI.⁴¹ Key aspects concerning human factors include:

1. Ratio of People to Robots: the ratio of people to robots directly affecting HRI in a system.⁴¹

2. Level of Shared Interaction Among Teams: the collaborative versus independent controls between human and robot agents.⁴¹

3. Interaction Roles: for example, the five roles that humans can take in an interaction with robots, as defined by Scholtz: supervisor, operator, teammate, mechanic/programmer, and bystander.⁴²

4. Type of Human–Robot Physical Proximity: the level of physical proximity between humans and robots. Huttenrauch and Eklundh proposed five modes: avoiding, passing, following, approaching, and touching.⁴³ In addition, the emerging subdomain under HRI named physical Human–Robot Interaction (pHRI) focuses on the physical side of the brain–body duality in robotic systems with the presence of the human. Research topics concerning pHRI are shown in Figure 2 by De Santis et al.

5. Decision Support for Operators: the interface between human and robot—specifically how information is conveyed to form decisions, e.g., how sensors are integrated.

Figure 2.

A “map” of robotics for anthropic domains: main issues and superposition for pHRI, illustrated by De Santis et al.³⁹ in 2008.

With the cyborg as a metaphor, human abilities have also been enhanced and extended by computational and robotic tools. People can now sense and visualize information that was previously out of human’s inherent capability with the aid of external instruments, such as smart devices (e.g., smart phone⁴⁴ and smartwatch⁴⁵), skin-based interface,^46,47 and wearable (e.g., head-mounted displays,^48,49 backpack and portable hand-held devices⁵⁰). In addition, hybrid or mixed reality starts to merge physical and digital environments into one, allowing humans to have more meaningful interactions and controls over computational and robotic tools.⁴⁴

Inspirations: Biology and game theory

Bioinspirations

Multi-agent robotics often draw inspirations from nature, such as self-assembly (e.g., crystals of bacterial flagella,⁵¹ multi-cellular organism development,^52,53 and species of ants, bees, and wasps⁵⁴). The collective behaviors in groups emerge from local interactions, with the overall pattern tuned by positive and negative feedback loops. The recent implementation of opinion vectors also allows for a smooth switch between consensus and dissensus among uniformed swarm group,⁵⁵ enabling more nuanced control over multi-robotic systems.

Through a control matrix, swarm algorithms allow multi-agent robots to be responsive and flexible, which are desirable qualities in forming safe and robust interactions between robots and humans. Goodrich et al.⁵⁶ explored the leader (attract) and predator (repel) styles of human influences through a bio-inspired robot team. According to Goodrich, ”leaders are more effective in influencing coherent flocks, but predators can be used to divide the flock into sub-flocks, yielding higher performance on some problems.⁵⁷” Their research combines HRI with Bio-Inspired Robot Teams (BIRT) into HuBIRT (human interaction with bio-inspired robot teams), providing a new way of considering humans’ relationship with swarm robots. Furthermore, the biological concepts have inspired new roles in robotic teams such as mediators⁵⁸ to facilitate a more integrative control system that can fulfill a wider range of spatio-temporal tasks.⁵⁹ Besides the branch of bio-inspired robotic models focusing on mediator’s roles, another branch discusses the different algorithms for swarm communication models (e.g., metric, topological, and visual) and emphasizes the importance of the deliberate selection of such models to enhance the task performance of a swarm system.⁶⁰

The bioinspirations for swarm robotics introduce structures and hierarchies into the management and coordination systems. Instead of the traditional controller-agent⁶¹ or peer relationship⁶² with robots, the biological examples mentioned above suggest roles in-between. As mediators, humans can coordinate a large group of swarm robots with a more flexible and fine-tuned control. However, most researches in this area remain in the theoretical or simulated stage, with few examples in real-life robotic applications. It is a great opportunity and challenge to bring these bio-inspired multi-robotic models into design and fabrication applications.

Economics and game theory

Economics, especially game theory, has also inspired control methods for collective robotic systems with concepts of negotiation and decision-making. Studies inspired by game theories, in general, focus on the cooperation of selfish rational agents⁷ in achieving utility-maximizing scenarios with respect to other agents’ decisions. In multi-robotic task allocation, one can construct a price-based task market, and each robot is programmed to act selfishly and buy⁶³ or auction^64,65 tasks for maximum profit.⁶⁶

Game theory is broadly applicable to various robotic problem types⁶⁷ and can yield efficient and responsive computational frameworks.⁶⁸ In 1993, LaValle and Hutchinson brought game theory into multi-robotic scenarios by conceptually applying game theory to topics such as “multiple robot coordination, high-level strategy planning, and information gathering.”⁶⁷ In 2008, Meng⁶⁸ demonstrated a dynamic-programming function, where a nonzero-sum game is formed so that optimal overall robotic performance can be accomplished. This game-theory-based framework is designed to be efficient and robust in a dynamic environment with real-time responsiveness.

Game theory has influenced not only multi-agents robotics but also human–robot interaction. In 2008, Lee and Hwang⁶⁹ proposed a gametheoretic approach to communicative interaction and cooperative decision-making. In 2016, Hadfield-Menell et al.⁷⁰ proposed cooperative inverse reinforcement learning (CIRL) that effectively achieves value alignment in a partial-information human–robot game. Game theory can also be used to better understand HRI. For example, in 2016, Wu and Paeng used the game theory approach to show that“a human may grow to trust a robot teammate more than a human teammate.”^71,72 Besides, studies in heterogeneous teams also draw inspirations from game theory. In 2010, Tarasenko⁷⁴ applied Lefebvre’s Reflexive Game Theory (RGT)⁷³ to model mixed groups of humans and robots, where both parties are represented with a unified hierarchy and methods of solution are considered simultaneously with practical applications.

Both the biological and economic concepts provide new inspirations for integrating humans into a heterogeneous robotic team with a less centralized and more flexible control system. This flexibility in coordination can help humans better predict the robots’ next moves, split tasks with robots in a more organic manner, and influence task executions on the fly. However, researches in the theses areas are currently conducted mainly in simulated or lab environments. It remains a challenge to transfer these methods into real-world applications, especially in the construction industry, efficiently and robustly.⁷⁵

Research Problems across CRC and HRI

Goodrich and Schultz defined HRI problems to be “to understand and shape the interactions between one or more humans and one or more robots” that can be broken down into the following constituent parts (44):

1. Level and behavior of autonomy

2. Nature of information exchange

3. Structure of the team

4. Adaptation, learning, and training of people and the robot

5. Shape of the task.

Meanwhile, as defined by Petersen et al.,² CRC consists of four major areas: Robotic Systems, Assembly Algorithms, Building Design, and Functional Material (Figure 1). Overall, the research scope of CRC and HRI is not just limited to technical challenges such as robotic systems and material manipulation methods but extends to areas of social dynamics and team management, such as in swarm behavior, task assignment, human–robot collaboration, and more. Although both CRC and HRI are already interdisciplinary fields, a meaningful bridge can be made between the two by carefully laying out related research problems onto research axes relevant to the robotic and architectural scope.

Research axes

In this paper, we propose two research axes (autonomy–collaboration and design–fabrication, see Figure 3) to situate the ranges of topics encountered in CRC and HRI, as well as industry-specific topics regarding human collectives in building sector. The autonomy–collaboration axis covers technical aspects in robotics such as sensing, controlling, and communicating. The design–fabrication axis helps situate the multi-agent robotic researches into the field of architecture. Together, these two axes aim to map out research domains and gaps in collective human–robot construction. In Figure 3, a gradient diagram qualitatively illustrates how sub-branches of CRC, HRI, and architectural design fabrication topics mentioned in the reference list of this review paper fit into the research scope spanning across autonomy–collaboration and design–fabrication axes. The main diagram in Figure 3 is formed by merging two human and robotic collective construction diagrams into one. As a result, the horizontal Autonomy–Collaboration–Autonomy axis in the key diagram is mirrored around Collaboration with the left and right ends illustrating each of the robotic and human collectives.

Figure 3.

Bridging the collectives: collective human–robot construction.

Autonomy–collaboration

Autonomy

Autonomy is a research topic that is critical to both CRC and HRI. Beer et al. have provided a loose definition for autonomy in humans or machines as “the extent to which a system can carry out its own processes and operations without external control.”⁷⁶ They also provided a more specific definition of robotic autonomy to be “the extent to which a robot can sense the environment, plan based on that environment, and act upon that environment, with the intent of reaching some goal (either given to or created by the robot) without external control.”⁷⁶

There are two major groups of opinions on autonomy’s implications in the HRI context. Huang et al. suggests that higher autonomy yields less interaction between humans and robots.^77–80 Their research on unmanned system (UMS) aims to maximize system performance with minimal human intervention. Goodrich and Olsen echo the point by associating the idea of neglect with autonomy. They argue that the length of a time period in which a person can neglect the robot is directly associated with how effective the robot performs its tasks independently.^81,82 Therefore, less interaction is required if the robot has higher level of autonomy.⁴¹

A contrasting perspective suggests that increasing the level of autonomy requires more complex types of collaboration among agents.^9,11,83,84 In other words, there’s a positive correlation between autonomy and collaboration. With higher levels of autonomy in robots, a more balanced partnership can be established, where humans and robots work together toward a common goal with proper communication and complementing skill sets.^8,84 For example, Fong suggested a model where robots can ask a human for advice and take advantage of human perception and cognition especially when facing arbitrary situations.^8,85

While robotic system can range from fully autonomous to teleoperated, adjustable autonomy has introduced the notion of switching between a range of autonomy modes or levels.⁸⁶ Adjustable autonomy opens up a design space which allows for fine-tuning the level of HRI to achieve desired performances. For example, adjustable autonomy allows for strategic direction of human attention, which can helpful when human focus is requires in design decision on-site or emergency response in hazardous conditions.

Different modes of autonomy have been proposed by existing robotic control frameworks. For example, Bruemmer et al. from the Idaho National Engineering and Environmental Laboratory (INEEL) proposed a robot control architecture that consists of four primary autonomy modes: teleoperation (human has full and continuous control of the robot at a low level), safe mode (human directs the robot’s movement; robot takes initiatives to protect itself), shared control (The robot takes the initiative to choose its own paths for local objectives), and full autonomous (robot performs global planning and only requires high-level tasking input from humans).⁸⁷ Bruemmer and Walton⁸⁸ also created a dynamic umbrella of adjustable autonomy control that can “support a spectrum of team interactions as individual capabilities change and needs arise.”

Collaboration

Collaboration requires agents to have a certain level of autonomy but focuses on the coordination between multiple agents. Wang identifies five team-performance measures: trust, compliance, transparency, mission success, and correct decisions, with trust being a foundation in human–robot teams.^89,90 Researchers have shown that a better understanding of a robot’s decision-making mechanisms can help enhance trust for human operators.⁹¹ In Hoffman’s latest fluency metric scale, he provided a list of subjective items to consider when evaluating Human–Robot Collaboration.^92,93 By analyzing through aspects like trust in the robot, human–robot fluency, positive teammate traits, and more, human–robot collaboration is studied through a more subjective and experience-based perspective.

Collaborations among humans are increasingly influenced by advancements in machines. Human–Computer Interaction (HCI) provides new interfaces for directed, collaborative, and passive crowd-sourcing in the context of collective intelligence.⁹⁴ Besides, the prevalence of cloud computing has inspired the concept of Human–Robot Cloud. HRC enables distributed computation, storage, sending, and actuator networks.⁹⁵ Meanwhile, researches related to physical collaboration among collectives are also essential. The study of Physical Human–Robot Interaction (pHRI) concerns HRI in close physical proximity. It explores human-friendly robot design (e.g., lightweight machines with flexible force control⁹⁶) and human-aware motion planning (e.g., with velocity constraints,⁹⁷ sampling-based costmap planners,⁹⁸ and dynamical systems⁹⁹). Aside from challenges regarding HRI in close proximity, another research lens focuses on embodying interactions in space. As stated by Eng et al., an interactive space consists of “data from human activity, human opinions, and the activity of the space.”¹⁰⁰ In these settings, changes in human groups influence the distributed manipulators to form an interactive environment that exhibits collective agency across human and machine. Examples of interactive spaces include the Flexing Room¹⁰¹ and Ada.¹⁰²

How humans and robots communicate with each other is important for collaboration. In 2015, Mavridis¹⁰³ gave a thorough review on the topic of verbal and non-verbal human–robot interactive communication. Mavridis used a list of 10 desiderata to guide his review:

1. Breaking the “simple commands only”¹⁰⁴ barrier.

2. Multiple speech acts.¹⁰⁵

3. Mixed initiative dialogue (e.g., the Karlsruhe Humanoid¹⁰⁶ and the Biron and Barthoc systems at Bielefeld¹⁰⁷).

4. Situated language and the symbol grounding problem (e.g., ELIZA,¹⁰⁸ the Loebner Prize,¹⁰⁹ and the POETICON project¹¹⁰).

5. Affective interaction (e.g., the Kismet robot¹¹¹ and the ATR Robovie robot¹¹²).

6. Motor correlates and Non-Verbal Communication (e.g., eye gaze cues^113,114 and gestures¹¹⁵).

7. Purposeful speech and planning (which concerns navigation through uncertainties).

8. Multi-level learning (e.g., Machine Learning¹¹⁶ and crowd-sourcing in HR¹¹⁷).

9. Utilization of online resources and services (e.g., the Roboearth project¹¹⁸ and its Rapyuta cloud engine¹¹⁹).

10. Miscellaneous abilities (that can deal with multiple conversational partners¹²⁰ with multilingual capabilities¹²¹).

Mavridis’ review suggests a trend toward more fluid communication, encompassing numerous formats and the dynamic integration among them. In robotic fabrication, enhanced communication among humans, robots, built structures, and the larger environment enables feedback loops that help to balance the fabrication and design concerns.¹²² Not only can we develop new ways of strengthening connectivity across the builder, tool, and environment, but we can also rethink built space as a medium, where the building becomes a living organism that can shift and adjust based on the inhabitants’ inputs. For example, in robotic research, Nagpal has promoted the idea of communicating through built structures via collective agents.¹³ In architectural theory, Kilian¹⁰¹ proposed the flexing room where the human–machine interface is represented by the space itself.

Design–fabrication

The design–fabrication axis aims to tie the research domain to the building sector. The design end of this axis is related to the virtual aspects, including topics in decision-making, information representation and exchange, communication, and more. Meanwhile, fabrication concerns the physical processes, including but not limited to material manipulation, localization, navigation, tolerances, collision avoidance, robotic kinematics, and human–robot physical interactions.

Topics in design draw references from fields such as Human–Computer Interaction¹²³ and how CAD and BIM influenced the building industry, but with additional consideration in collective robotics as the driving factor. Bock first proposed the term Robot-Oriented Design (ROD) in 1988³⁶ and published a same-titled book in 2015.³⁷ According to Bock, ROD is concerned with “the co-adaptation of construction products and automated or robotic technology, so that the use of such technology becomes applicable, simpler, more robust, or more productive or efficient.” An example such as ROD lies in the middle of the design–fabrication axis. Besides, emerging research starts to explore how design tasks can be split and distributed to robotic collectives, where individual robots make decisions based on inputs from the other robots and the environment. For example, Yablonina’s co-design strategy with larger groups of smaller robots showcases a codependent development of the process, object, and machine.¹²⁴

Discussion

Collective human–robot construction

After reviewing CRC and HRI, we identify a potential research domain situated in between the two and define it as follows:

Collective Human–Robot Construction (CHRC) concerns multi-agent construction involving both human and robotic collectives. It is an emerging interdisciplinary field that combines collective fabrication, human–robot interaction, and heterogeneous teams. Research focused on CHRC spans from autonomy to collaboration, indicating novel ways of designing and fabricating.

Motivations for Defining CHRC Research Space

CHRC is a unique field compared to its neighboring ones. It differs

1. from CRC with a focus on human factors,

2. from HRI due to the emphasis on collectiveness, and

3. from human–robot teams through its application in construction.

The purpose of this review is to draw attention to the specific overlapping research topics concerning collectiveness, human–robot teams, and construction, and provide the terminology of CHRC to refer to this particular area.

CHRC is critical in joining forces and intelligence of robotic and human collectives to achieve greater impact in the field of construction (ranging from team dynamics¹²⁵ to economic profits¹²⁶). The distributed timber construction with collective robots acting as joints¹⁹ is one recent example among others (e.g., tensile web construction¹⁴ and micro-robot truss construction¹⁵) that shows the potential of CRC in creating complex spatial structures. Both the emergence of dynamic autonomous control and diverse collaboration models have opened up possibilities for new ways of designing. For example, Yablonina’s mobile robotic fabrication system for filament structures takes a step further from ROD to imply a co-design scenario between human and distributed robotic agents.¹²⁴ From a broader industrial point of view, distributed collective systems have the advantage of responsiveness and robustness.^55,127,128 Combining human sensitivity and creativity with robotic strength and accuracy can help solve existing problems concerning tolerance and scalability¹²⁹ to accelerate the transferring of novel fabrication methods from lab to real-world construction sites. Therefore, in the long run, CHRC has the potential to help expand the adoption of robotic technology especially for on-site construction challenges.^130,131

In a nutshell, CHRC is a promising field for exploring:

1. New formal expressions

2. New ways of designing and fabricating

3. Expanding the scale of robotic application in construction sector

4. Increasing efficiency in communication, computation, task allocation, and material usage.

Future challenges and opportunities in CHRC

The emerging CHRC field is filled with challenges and opportunities. Moving forward, we need to not only tackle technical problems (e.g., multi-agent decision-making, control, communication, and learning) but also design integrative systems to efficiently bring the two collectives together (humans and robots) in pursuing design and construction tasks in real-world practice. While there is rich literature in conceptualizing collective systems and human factors, further work is needed to synthesize the two themes into construction applications.

Figure 3 provides a visual indication of potential research gaps to fill in CHRC. The clustering of robot-related topics toward the fabrication end of the axis indicates a potential utilization of robots in assisting design processes. The blue column at the center gathers topics directly related to human and robot collaboration. Thus, there is rich potential in bridging and blending such topics. For example, social robots can be paired with robot-aware behaviors to create design fabrication scenarios where robots and humans are both orientating their decisions and behaviors based on the other’s characteristics. Another example is to connect Bio-Inspired Control and Interaction Roles in swarm robotic teams (Figure 3). The topic has already been slightly touched upon in Goodrich’s HuBIRT research,⁵⁶ although it was not oriented toward architectural fabrication. Similarly, due to the mirrored layout along the Autonomy–Collaboration–Autonomy axis in Figure 3, one can also fold the diagram vertically in half and find potential research bridges across the “robots” and “humans” areas on symmetric locations in the diagram. Such examples include the symmetry between robot’s self-healing and human-operated architectural maintenance, between robots’ collective perception and human’s extended connectivity through smart devices, or between robot sensing and smart wearable. It is also worth mentioning that the research topics that are qualitatively located on Figure 3 are dynamically shifting with on-going development of the subjects. Therefore, there can be motivations to actively push one topic toward a certain research direction. For example, 3D Printing is currently located toward the robot’s Autonomy end of the axis. However, this does not mean that 3D Printing would not benefit from more feedback and input from sensors and human operators.

To accelerate the bridging across the disciplines, iterative prototypes of collective construction systems can be developed to further expand humans’ roles in existing CRC systems or collective behaviors in HRI. For example, Petersen et al. identified the human element in CRC by stating that “Future research will also reveal where humans are best placed with respect to CRC, e.g., to define goal specifications, to provide online corrections when robots commit rare, but inevitable errors, or to support and maintain robots.”² Besides, from the perspective of HRI, more complex, nuanced, and flexible interaction among heterogeneous multi-robot–human teams shall be further developed. Based on this review paper and its corresponding diagram (Figure 3), researchers may further advance CHRC through:

1. Bringing heterogeneous team inspirations (e.g., from the inter-species relationship in nature and game theories) into collective construction systems involving diverse human roles and robotic types;

2. Iteratively transferring collective construction systems from a simulated or controlled lab environment into real-world construction settings and building environment,¹²⁹ and this requires more considerations of problems such as latency, tolerance, control, and management;

3. Creating proof-of-concept projects to set cornerstones in adapting techniques and methods from the neighboring field (e.g., HRI, CRC) into the discipline of architecture and design;

4. Leveraging fluid and dynamic control and communication models^88,95,99 to distribute design decisions (e.g., co-design strategy¹²⁴) and form complex structural and material systems;^13,132

5. Developing management strategies to prepare for a safe, efficient, and flexible application of CHRC on site;

6. Investigating topics of group creativity and social dynamics in heterogeneous human–robot teams. For example, one can imagine scenarios where humans improvise with robots in addition to pre-composed assembly procedures to achieve site-specific structures. Alternatively, one can also utilize human–robot group creativity for emergency management at the construction site;

7. Rethinking the roles of design and construction professionals, specifically in how to best allocate humans’ capabilities and attention, given the potentials and constraints of robotic intelligence and kinematics (e.g., ROD^36,37); and

8. Exploring new forms and functionalities of space, providing dual dimensions in physical structure and embodied control. For example, smart devices and wearable can be designed to empower human body with higher sensibility and capability when collaborating with robots. Thus, more fabrication strategies using mixed or hybrid reality can be developed to provide additional platforms for humans, computers, and robots to create and construct collectively.

Conclusions

Both robotic systems and human groups in the construction sector observe a trend toward decentralization and collaboration. This review’s purpose is to identify gaps between Collective Robotic Construction (CRC) and Human–Robot Interaction (HRI) literature to define a new research domain in Collective Human–Robot Construction (CHRC). Both CRC and HRI are young but active research domains, and there are on-going discussions on how humans can interact with collective robotic systems. However, there is a less-touched research space in reviewing human collectives (e.g., institutions, professions, corporations) in relation to multi-agent robotics, as well as one targeted toward the construction industry connecting collective robotic tools with the existing operational and informational structure such as BIM. This paper reveals promising research areas in bridging the collectives (humans and robots, see Figure 3). Developing CHRC is essential for leveraging robotic technology within the existing building industry to create scalable impact. By combining humans and robots’ collective intelligence and capabilities, we can increase efficiencies (in communication, material processing, and fabrication) and create formal and organizational structures that are far more complex than either party can create alone. Thus, the emphasis on human elements in robotic construction reveals potential for a safe and pleasant collaboration model where human creativity and sensitivity can be best manifested. By bringing these two seemingly disparate research areas together into the field of construction, we can open up previously unknown possibilities for architecture and design.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Isla Xi Han

Stefana Parascho

References

Willmann

Block

Hutter

, et al. Robotic fabrication in architecture, art and design 2018: foreword by Sigrid Brell-Cokcan and Johannes Braumann, association for robots in architecture. Berlin, Germany: Springer, 2018.

Petersen

Napp

Stuart-Smith

, et al. A review of collective robotic construction. Sci Robotics 2019; 4: 28.

Felbrich

Prado

Saffarian

, et al. Multi- machine fabrication: an integrative design process utilising an autonomous uav and industrial robots for the fabrication of long-span composite structures. Cambridge, MA: Massachusetts Institute of Technology, 2017.

Parascho

Gandia

Mirjan

, et al. Cooperative fabrication of spatial metal structures. Fabricate 2017; 24–29.

Rust

Jenny

Gramazio

, et al. Spatial wire cutting: cooperative robotic cutting of non-ruled surface geometries for bespoke building components. In: Chien

Choo

Schnabel

, et al. (eds). Living Systems and Micro-Utopias: Towards Continuous Designing, Proceedings of the 21st International Conference of the Association for Computer-Aided Architectural Design Research in Asia CAADRIA 2016. Ann Arbor, MI: Hong Kong: Association for Computer Aided Design in Architecture ACADIA; 2016, pp. 529–538.

Wagner

Alvarez

Groenewolt

, et al. Towards digital automation flexibility in large- scale timber construction: integrative robotic prefabrication and co-design of the buga wood pavilion. Construction Robotics 2020; 4: 3187–3204.

Cao

Fukunaga

Kahng

. Cooperative mobile robotics: antecedents and directions. Autonomous Robots 1997; 4(pp. 7–27): 1.

Fong

Thorpe

Baur

. Collaboration, dialogue, human-robot interaction. In: Robotics research. Berlin, Germany: Springer; 2003, pp. 255–266.

Goodrich

Schultz

. Human-robot interaction: a survey. Boston, MA: Now Publishers Inc, 2008.

10.

Steinfeld

Fong

Kaber

Lewis

Scholtz

Schultz

Goodrich

Common metrics for human-robot interaction. In Proceedings of the 1st ACM SIGCHI/SIGART conference on Human-robot interaction, Salt Lake City, UT, 2–3 March 2006. pp. 33–40, 2006.

11.

Thrun

. Toward a framework for human-robot interaction. Human–Computer Interaction 2004; 19(1–2), pp. 9–24.

12.

Saidi

Bock

Georgoulas

. Robotics in construction. In: Springer handbook of robotics. Berlin, Germany: Springer; 2016, pp. 1493–1520.

13.

Werfel

Petersen

Nagpal

. Designing collective behavior in a termite-inspired robot construction team. Science (New York, N.Y.) 2014; 343(6172), pp. 754–758.

14.

Braithwaite

Alhinai

Haas-Heger

, et al. Tensile web construction and perching with nano aerial vehicles. In: Robotics Research. Berlin, Germany: Springer; 2018, pp. 71–88.

15.

Hsu

Wong-Foy

Mccoy

, et al. Application of micro-robots for building carbon fiber trusses. In: 2016 International Conference on Manipulation, Automation and Robotics at Small Scales (MARSS), Paris, France, 18–22 July 2016. IEEE, 2016, pp. 1–6.

16.

Sheridan

. Human-robot interaction. Hum Factors J Hum Factors Ergon Soc 2016; 58(4), pp. 525–532.

17.

Ferber

Weiss

. Multi-agent systems: an introduction to distributed artificial intelligence, vol. 1. Boston, MA: Addison-Wesley Reading, 1999.

18.

Andreen

Jenning

Napp

, et al. Emergent structures assembled by large swarms of simple robots. In: Velikov

Manninger

del Campo

, et al. (eds) Acadia 2016 Posthuman frontiers: Data, Designers, and Cognitive Machines. Ann Arbor, MI: Association for Computer Aided Design in Architecture ACADIA; 2016, pp. 54–61.

19.

Weber

Wood

Bucklin

, et al. Distributed robotic timber construction. Abingdon, UK: Routledge, 2019.

20.

Nguyen

. Intelligent collectives: theory, applications, and research challenges. Cybernetics Syst 2018; 49(5–6): 261–279.

21.

Sawyer

. Improvisational creativity: an analysis of jazz performance. Creativ Res J 1992; 5(3): pp. 253–263.

22.

Sawyer

. Group creativity: music, theater, collaboration. In: Collective ideation: creativity, teamwork, and collaboration. East Sussex, UK: Psychology Press; 2014, pp. 262–304.

23.

Rickards

Runco

Moger

(eds) The Routledge companion to creativity. Abingdon, UK: Routledge; 2008.

24.

Weinberg

Blosser

Mallikarjuna

, et al. The creation of a multi-human, multi-robot interactive jam session. In NIME, Pittsburgh, PA, 4–6 June 2009. pp. 70–73.

25.

Savery

. Machine learning driven musical improvisation for mechanomorphic human-robot interaction. In Companion of the 2021 ACM/IEEE International Conference on Human-Robot Interaction, Boulder, CO, 8–11 March 2021. pp. 559–561, 2021.

26.

Mikalauskas

Wun

, et al. Improvising with an Audience-Controlled Robot Performer. In Proceedings of the 2018 Designing Interactive Systems Conference, Hong Kong, 9–13 June 2018. pp. 657–666, 2018.

27.

Landemore

Elster

. Collective wisdom: Principles and mechanisms. Cambridge, UK: Cambridge University Press, 2012.

28.

Peters

Heraud

. Toward a political theory of social innovation: collective intelligence and the co-creation of social goods. J Self-Governance Manag Econ 2015; 3(3).

29.

Wooldridge

. An introduction to multiagent systems. Hoboken, NJ: John Wiley Sons, 2009.

30.

Virdhagriswaran

Osisek

O’connor

. Standardizing agent technology. Stand View 1995; 3(3), pp. 96–101.

31.

Chiariglione

. Fipa 97 specification. Washington, DC: Foundation for Intelligent Physical Agents, 1997.

32.

Bradshaw

Dutfield

Benoit

, et al. Kaos: toward an industrial-strength open agent architecture. Softw Agents 1997; 13: pp. 375–418.

33.

Flores-Mendez

. Towards a standardization of multi-agent system framework. XRDS: Crossroads, ACM Mag Students 1999; 5(4): 18–24.

34.

White

. Mobile agents. In: Bradshaw

(ed) Software Agents. Menlo Park, CA: American Association for Artificial Intelligence; 1997.

35.

Autodesk . What is BIM: Building information modeling. San Rafael, CA: Autodesk, https://www.autodesk.com/solutions/bim (Accessed 15 Dec 2020).

36.

Bock

. “Robot oriented design”, architectural product engineering. London, UK: seko, 1988.

37.

Bock

Linner

. Robot oriented design. Cambridge, UK: Cambridge university press, 2015.

38.

Bock

Linner

. Integration along the value chain in construction through robot oriented management. Arch Manag Digital Arena 2011; 21.

39.

De Santis

Siciliano

De Luca

, et al. An atlas of physical human-robot interaction. Mechanism Machine Theor 2008; 43(3): 253–270.

40.

Tversky

Kahneman

. Judgment under uncertainty: heuristics and biases. Science 1974; 185(4157), pp. 1124–1131.

41.

Yanco

Drury

. Classifying human-robot interaction: an updated taxonomy. In: 2004 IEEE International Conference on Systems, Man and Cybernetics (IEEE Cat. No. 04CH37583), The Hague, Netherlands, 10–13 October 2004, 3. IEEE, 2004, pp. 2841–2846.

42.

Scholtz

. Theory and evaluation of human robot interactions. In: 36th Annual Hawaii International Conference on System Sciences, Big Island, HI, 6–9 January 2003. IEEE, 2003, p. 10.

43.

Huttenrauch

Eklundh

Investigating socially interactive robots that give the right cues and make their presence felt. In Proceedings of the CHI 2004 Workshop on Shaping Human-Robot Interaction, Vienna, Austria, 24–29 April 2004. http://www.bartneck.de/workshop/chi2004, pp. 17–20, 2004.

44.

Atanasova

Mitterberger

Sandy

, et al. Prototype as artefact - design tool for open-ended collaborative assembly processes. Bar Harbor, ME: ACADIA 2020 Distributed Proximities, 2020; pp. 350–359.

45.

Vasey

Grossman

Kerrick

, et al. The hive: a human and robot collaborative building process. In ACM SIGGRAPH 2016 Talks, Anaheim, CA, 24–28 July 2016. pp. 1–2, 2016.

46.

Gannon

. Human-centered Interfaces for autonomous fabrication machines. (Doctoral Dissertation). Pittsburgh, PA: Carnegie Mellon University), 2018.

47.

Harrison

Tan

Morris

. Skinput: appropriating the body as an input surface. In Proceedings of the SIGCHI conference on human factors in computing systems, Denver, CO, 7–11 May 2010. pp. 453–462, 2010.

48.

Abe

Hotta

, et al. Digital Construction - Demonstration of Interactive Assembly Using Smart Discrete Papers with RFID and AR codes. In: Janssen

Loh

Raonic

et al. (eds). Protocols, flows, and glitches - Proceedings of the 22nd CAADRIA Conference, CUMINCAD. Suzhou, China: Xi’an Jiaotong-Liverpool University; 2017, pp. 75–84.

49.

Newnham

van den Berg

Beanland

. Making in mixed reality. In: Proceedings of ACADIA. Bar Harbor, ME: Acadia Publishing Company; 2018, pp. 88–97.

50.

Mitterberger

Dörfler

Sandy

, et al. Augmented bricklaying. Construction Robotics 2020; 4(3): pp. 151–161.

51.

Whitesides

Grzybowski

. Self-assembly at all scales. Science (New York, N.Y.) 2002; 295(5564), pp. 2418–2421.

52.

Camazine

Deneubourg

J-L

Franks

, et al. Self- organization in biological systems. Princeton, NJ: Princeton university press, 2003.

53.

Lawrence

. The making of a fly: the genetics of animal design. Oxford, UK: Blackwell Scientific Publications Ltd, 1992.

54.

Anderson

Theraulaz

Deneubourg

J-L

. Self-assemblages in insect societies. Insectes Sociaux 2002; 49(2): pp. 99–110.

55.

Bizyaeva

Franci

Leonard

. A general model of opinion dynamics with tunable sensitivity. arXiv 2020 (preprint arXiv:2009.04332 [math.OC.].

56.

Goodrich

Pendleton

Sujit

, et al. Toward human interaction with bio-inspired robot teams. In: 2011 IEEE International Conference on Systems, Man, and Cybernetics, San Diego, CA, 9–12 October 2011. IEEE, 2011, pp. 2859–2864.

57.

Goodrich

Pendleton

Kerman

, et al.

What types of interactions do bio-inspired robot swarms and flocks afford a human?

Proc Robot: Sci Syst 2013; VIII: 105–112.

58.

Jung

S-Y

Goodrich

. Multi-robot perimeter- shaping through mediator-based swarm control. In: 2013 16th International Conference on Advanced Robotics (ICAR), Montevideo, Uruguay, 25–29 November 2013. IEEE, 2013, pp. 1–6.

59.

Brown

Jung

S-Y

Goodrich

. Balancing human and inter-agent influences for shared control of bio-inspired collectives. In: 2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC), San Diego, CA, 5–8 Oct 2014. IEEE, 2014, pp. 4123–4128.

60.

Haque

Baker

Ren

, et al. Analysis of biologically inspired swarm communication models. In: Advances in Hybridization of Intelligent Methods. Berlin, Germany: Springer; 2018, pp. 17–38.

61.

Ambrose

Askew

Bluethmann

, et al. The development of the robonaut system for space operations. In: Proceedings of ICAR, Invited Session on Space Robotics, Nitra, Slovak Republic, 26–27 June 2001. 2001.

62.

Fong

Nourbakhsh

. Interaction challenges in human-robot space exploration. Interactions 2005; 12(2), pp. 42–45.

63.

Shapley

Shubik

. The assignment game i: the core. Int J Game Theor 1971; 1(1): pp. 111–130.

64.

Bertsekas

. The auction algorithm for assignment and other network flow problems: a tutorial. Interfaces 1990; 20(4): pp. 133–149.

65.

Mcafee

Mcmillan

. Auctions and bidding. J Econ Lit 1987; 25(2): pp. 699–738.

66.

Gerkey

. On multi-robot task allocation. Los Angeles, CA: University of Southern California, 2003.

67.

Lavalle

Hutchinson

. Game theory as a unifying structure for a variety of robot tasks. In: Proceedings of 8th IEEE International Symposium on Intelligent Control, Chicago, IL, 25–27 August 1993. IEEE, 1993, pp. 429–434.

68.

Meng

. Multi-robot searching using game-theory based approach. Int J Adv Robotic Syst 2008; 5(4), p. 44.

69.

Lee

Hwang

J-H

. Human-Robot Interaction as a Cooperative Game. In: Trends in Intelligent Systems and Computer Engineering. Berlin, Germany: Springer; 2008, pp. 91–103.

70.

Hadfield-Menell

Russell

Abbeel

, et al. Cooperative inverse reinforcement learning. In Advances in neural information processing systems, Barcelona, Spain, 5–10 December 2016. pp. 3909–3917, 2016.

71.

Paeng

Boerkoel

Human-robot trust and cooperation through a game theoretic framework. In AAAI, Phoenix, AZ, 12–17 February 2016. pp. 4246–4247, 2016.

72.

Paeng

Linder

, et al. Trust and cooperation in human-robot decision making. In 2016 AAAI fall symposium series, Arlington, Virginia, 17–19 November 2016. 2016.

73.

Lefebvre

. Lectures on the Reflexive Games Theory. Morrisville, NC: Lulu. com, 2010.

74.

Tarasenko

. The inverse task of the reflexive game theory: theoretical matters, practical applications and relationship with other issues. arXiv 2010, preprint arXiv:1011.3397.

75.

Wareham

. Designing robot teams for distributed construction, repair, and maintenance. ACM Trans Autonomous Adaptive Syst 2019; 14(1), pp. 1–29.

76.

Beer

Fisk

Rogers

. Toward a framework for levels of robot autonomy in human-robot interaction. J Hum Robot Interact 2014; 3(2): pp. 74.

77.

Huang

H-M

Messina

Wade

, et al. Autonomy measures for robots. In ASME International Mechanical Engineering Congress and Exposition, Anaheim, CA, 13–19 November 2004. vol. 47063, pp. 1241–1247, 2004.

78.

Huang

H-M

Pavek

Albus

, et al. Autonomy levels for unmanned systems (alfus) framework: An update. In: Unmanned Ground Vehicle Technology VII, vol. 5804. Washington, DC: International Society for Optics and Photonics; 2005, pp. 439–448.

79.

Huang

H-M

Pavek

Novak

, et al. A framework for autonomy levels for unmanned systems (alfus). Proc AUVSI’s Unmanned Syst North America 2005: 849–863.

80.

Huang

H-M

Pavek

Ragon

, et al. Characterizing unmanned system autonomy: contextual autonomous capability and level of autonomy analyses. In: Unmanned Systems Technology IX, vol. 6561. Washington, DC: International Society for Optics and Photonics; 2007, p. 65611N.

81.

Goodrich

Olsen

. Seven principles of efficient human robot interaction. In: SMC’03 Conference Proceedings. 2003 IEEE International Conference on Systems, Man and Cybernetics. Conference Theme-System Security and Assurance (Cat. No. 03CH37483), Washington, DC, 8 October 2003, 4. IEEE, 2003, pp. 3942–3948.

82.

Olsen

Goodrich

. Metrics for evaluating human-robot interactions. Proc PERMIS 2003; Vol. 2003, p. 4.

83.

Feil-Seifer

Skinner

Matarić

. Benchmarks for evaluating socially assistive robotics. Interaction Stud 2007; 8(3): pp. 423–439.

84.

Hoffman

Breazeal

. Collaboration in human- robot teams. In AIAA 1st Intelligent Systems Technical Conference, Chicago, IL, 20–24 September 2004. p. 6434, 2004.

85.

Fong

Thorpe

Baur

. Robot, asker of questions. Robotics Autonomous Systems 2003; 42(3–4): pp. 235–243.

86.

Baker

Yanco

. Autonomy mode suggestions for improving human-robot interaction. In: 2004 IEEE International Conference on Systems, Man and Cybernetics (IEEE Cat. No. 04CH37583), The Hague, Netherlands, 10–13 October 2004, 3. IEEE, 2004, pp. 2948–2953.

87.

Bruemmer

Dudenhoeffer

Marble

. Dynamic-autonomy for urban search and rescue. In: AAAI mobile robot competition. AAAI. Alberta, Canada; 2002, pp. 33–37.

88.

Bruemmer

Walton

. Collaborative tools for mixed teams of humans and robots. Tech. rep. Idaho Falls, ID: Idaho National Engineering And Environmental Lab Idaho Falls, 2003.

89.

Freedy

Devisser

Weltman

, et al. Measurement of trust in human-robot collaboration. In: 2007 International Symposium on Collaborative Technolo- gies and Systems, Orlando, FL, 21–25 May 2007. IEEE, 2007, pp. 106–114.

90.

Wang

Pynadath

Hill

. Trust calibration within a human-robot team: Comparing automat- ically generated explanations. In: 2016 11th ACM/IEEE International Conference on Human-Robot Interaction (HRI), Christchurch, New Zealand, 7–10 March 2016. IEEE, 2016, pp. 109–116.

91.

Lee

Moray

. Trust, control strategies and allocation of function in human-machine systems. Ergonomics 1992; 35(10): pp. 1243–1270.

92.

Hoffman

. Evaluating Fluency in Human-Robot Collaboration. IEEE Trans Human-Machine Syst 2019; 49(3): pp. 209–218.

93.

Hoffman

Breazeal

. Effects of anticipatory perceptual simulation on practiced human-robot tasks. Autonomous Robots 2010; 28(4): pp. 403–423.

94.

Bigham

Bernstein

Adar

. Human- computer interaction and collective intelligence. Handbook of Collective Intelligence 2015; 57.

95.

Mavridis

Bourlai

Ognibene

. The human-robot cloud: Situated collective intelligence on demand. In: 2012 IEEE international conference on cyber technology in automation, control, and intelligent systems (CY, BER), Bangkok, Thailand, 27–31 May 2012. IEEE, 2012, pp. 360–365.

96.

Haddadin

Croft

. Physical Human-Robot Interaction. In: Springer handbook of Robotics. Berlin, Germany: Springer; 2016, pp. 1835–1874.

97.

Shi

Collins

Jr Goldiez

, et al. Human-aware robot motion planning with velocity constraints. In: 2008 International Symposium on Collaborative Technologies and Systems, Irvine, CA, 19–23 May 2008. IEEE, 2008, pp. 490–497.

98.

Mainprice

Sisbot

Jaillet

, et al. Planning human-aware motions using a sampling-based costmap planner. In: 2011 IEEE International Conference on Robotics and Automation. Piscataway, NJ: IEEE; 2011, pp. 5012–5017.

99.

Saveriano

Hirt

Lee

. Human-aware motion reshaping using dynamical systems. Pattern Recognition Lett 2017; 99, pp. 96–104.

100.

Eng

Mintz

Verschure

. Collective human behavior in interactive spaces. In: Proceedings of the 2005 IEEE International Conference on Robotics and Automation, IEEE, 2005, pp. 2057–2062.

101.

Kilian

The flexing room architectural robot. An actuated active-bending robotic structure using human feedback, In: ACADIA, Mexico city, Mexico, 18–20 October 2018.

102.

Eng

Babler

Bernardet

, et al. Bernardet U, et al. Ada-Intelligent space: An artificial creature for the SwissExpo. 02. In: 2003 IEEE International Conference on Robotics and Automation (Cat. No. 03CH37422) IEEE, 2003, Vol. 3, pp. 4154–4159.

103.

Mavridis

. A review of verbal and non-verbal human-robot interactive communication. Robotics Autonomous Syst 2015; 63: pp. 22–35.

104.

Mavridis

Roy

. Grounded situation models for robots: Where words and percepts meet. In: 2006 IEEE/RSJ international conference on intelligent robots and systems, Beijing, China, 9–13 October 2006. IEEE, 2006, pp. 4690–4697.

105.

Searle

. A taxonomy of illocutionary acts. Minneapolis, MN: University of Minnesota Press, 1975.

106.

Stiefelhagen

Ekenel

Fugen

, et al. Enabling multimodal human-robot interaction for the Karlsruhe humanoid robot. IEEE Trans Robotics 2007; 23(5): pp. 840–851.

107.

Wrede

Buschkaemper

Muhl

, et al. Analyses of feedback in HRI. How People Talk Comput Robots, Other Artif Commun Partners 2006; p. 38.

108.

Weizenbaum

. Eliza—a computer program for the study of natural language communication between man and machine. Commun ACM 1966; 9(1): pp. 36–45.

109.

Mauldin

. Chatterbots, tinymuds, and the turing test: entering the loebner prize competition. AAAI 1994; 94: pp. 16–21.

110.

Wallraven

Schultze

Mohler

, et al. The poeticon enacted scenario corpus—a tool for human and computational experiments on action understanding. In: Face and gesture 2011. Piscataway, NJ: IEEE; 2011, pp. 484–491.

111.

Breazeal

Vela’ Squez

. Toward teaching a robot ‘infant’using emotive communication acts. In: Proceedings of the 1998 Simulated Adaptive Behavior Workshop on Socially Situated Intelligence, Zurich Switzerland. State College, PA: Citeseer, 1998, pp. 25–40.

112.

Komatani

Ito

Kawahara

, et al. Recognition of emotional states in spoken dialogue with a robot, Innovations in Applied Artificial Intelligence. In: International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, Amsterdam, The Netherlands, 17–21 June 2013. Berlin, Germany: Springer, 2004, pp. 413–423.

113.

Brennan

Chen

Dickinson

, et al. Coordinating cognition: the costs and benefits of shared gaze during collaborative search. Cognition 2008; 106(3): pp. 1465–1477.

114.

Fussell

Kraut

Siegel

Coordination of communication: fffects of shared visual context on collaborative work. In Proceedings of the 2000 ACM conference on Computer supported cooperative work, Philadelphia, PA, pp. 21–30, 2000.

115.

Waldherr

Romero

Thrun

. A gesture based interface for human-robot interaction. Autonomous Robots 2000; 9(2), pp. 151–173.

116.

Klingspor

Demiris

Kaiser

. Human- robot communication and machine learning. Appl Artif Intelligence 1997; 11(7): pp. 719–746.

117.

Depalma

Chernova

Breazeal

. Leveraging online virtual agents to crowdsource human- robot interaction. In: Proceedings of CHI Workshop on Crowdsourcing and Human Computation, Vancouver,Canada, 7–12 May 2011. New York, NY: ACM, 2011, pp. 53–56.

118.

Waibel

Beetz

Civera

, et al. Roboearth. IEEE Robotics Automation Mag 2011; 18(2), pp. 69–82.

119.

Hunziker

Gajamohan

Waibel

, et al. Rapyuta: the roboearth cloud engine. In: 2013 IEEE international conference on robotics and automation, Karlsruhe, Germany, 6–10 May 2013. IEEE, 2013, pp. 438–444.

120.

Matsusaka

Fujie

Kobayashi

Modeling of conversational strategy for the robot participating in the group conversation. In Seventh European conference on speech communication and technology, Aalborg, Denmark, 3–4 September, 2001.

121.

Holzapfel

. Towards development of multilingual spoken dialogue systems. In: Proceedings of the 2nd Language and Technology Conference, Tallin, Estonia, 4–5 April 2005. Citeseer, 2005.

122.

Snooks

Jahn

. Closeness: On the Relationship of Multi-agent Algorithms and Robotic Fabrication. In: Robotic fabrication in architecture, art and design 2016. Cham, Switzerland: Springer; 2016, pp. 218–229.

123.

Preece

Rogers

Sharp

, et al. Human-computer interaction. Addison-Wesley Longman Ltd., Edinburgh Gate Harlow, UK,1994.

124.

Yablonina

Menges

. Distributed fabrication: Cooperative making with larger groups of smaller machines. Architectural Design 2019; 89(2): pp. 62–69.

125.

Jung

Difranzo

Stoll

, et al. Robot assisted tower construction-a resource distribution task to study human-robot collaboration and interaction with groups of people. arXiv 2018 (preprint arXiv:1812.09548.

126.

Bauer

Wollherr

Buss

. Human-robot Collaboration: a Survey. Int J Humanoid Robotics 2008; 05(01): pp. 47–66.

127.

Petersen

Nagpal

WerfelTermes

. An autonomous robotic system for three-dimensional collective construction. Robotics: science and systems VII. Piscataway, NJ: IEEE, 2011.

128.

Werfel

Nagpal

. Extended stigmergy in collective construction. IEEE Intell Syst 2006; 21(2): pp. 20–28.

129.

Han

Bruun

EPG

Marsh

, et al. From concept to construction: a transferable design and robotic fabrication method for a building-scale vault. In: 40th Annual Conference of the Association for Computer Aided Design in Architecture: Distributed Proximities, ACADIA 2020, ACADIA, 2020, pp. 614–623.

130.

Keating

Leland

Cai

, et al. Toward site-specific and self-sufficient robotic fabrication on architectural scales. Sci Robotics 2017; 2(5).

131.

Melenbrink

Kassabian

Menges

, et al. Towards force-aware robot collectives for on-site construction. Proc Catalog 37th Annu Conf Assoc Comput Aided Des Architecture (ACADIA) 2017: pp. 382–391.

132.

Allwright

Bhalla

El-Faham

, et al. SRoCS: leveraging stigmergy on a multi-robot construction platform for unknown environments. In: International Conference on Swarm Intelligence, Hefei, China, 17–20 October 2014. Berlin, Germany: Springer, 2014, pp. 158–169.

Bridging the collectives: A review of collective human–robot construction

Abstract

Keywords

Introduction

Method

Review

Collectiveness

Human factors

Inspirations: Biology and game theory

Bioinspirations

Economics and game theory

Research Problems across CRC and HRI

Research axes

Autonomy–collaboration

Autonomy

Collaboration

Design–fabrication

Discussion

Collective human–robot construction

Motivations for Defining CHRC Research Space

Future challenges and opportunities in CHRC

Conclusions

Footnotes

Declaration of conflicting interests

Funding

ORCID iDs

References