1 Introduction

The globalization of economic activity is perhaps the defining trend of our time. It is reshaping not only the grand, macro level aspects of economic life but the personal aspects as well, including where, when, how, and with whom we perform our daily work (O’Hara-Devereaux and Johansen 1994).

The increased connectivity provided by the Internet in the modern world is unprecedented and allows people to collaborate through diverse media and communication channels. But despite all the new information and collaboration technologies (ICT), distributed teams still face many challenges that face-to-face teams do not. This means teaching students, i.e., the next-generation or practitioners, how to collaborate globally across disciplines, time, space, and culture is of increasing importance as distributed work becomes more prevalent and advantageous.

To achieve this objective, we have launched in 1993 the architecture, engineering, construction (AEC) Global Teamwork course established by the PBL Lab at Stanford in collaboration with universities worldwide. The AEC Global Teamwork course was described in previous papers. (Fruchter 1999, 2006; Fruchter and Emery 1999). AEC project team members determine the role of discipline-specific knowledge in a cross-disciplinary project-centered environment. It is through cross-disciplinary interaction that the team becomes a community of practitioners. The mastery of knowledge and skill requires individuals to move toward full participation in the sociocultural practices of a larger AEC community. The negotiation of language and culture is equally important to the learning process—through participation in a community of AEC practitioners; the students are learning how to create discourse that requires constructing meanings of concepts and uses of skills (Dewey 1928, 1958; Greeno 1998; Lave and Wenger 1991; Wenger 1998).

Key to this process is for each team member to build an awareness, appreciation, and understanding of the other disciplines. Awareness is identified when team members are knowledgeable of the other disciplines’ goals, objectives, and constraints. Appreciation is identified when team members can articulate questions and explanations that are meaningful across disciplines (Fruchter and Emery 1999). Understanding is identified when team members use concepts and terms from the other disciplines as they engage in a dialogue to build common ground and act together to produce a high-quality product.

We explore in this paper the relation between activities, communication channels and media, and common ground building in global teams. We used AEC Global Teamwork course offered in 2008–2009 as the testbed for our study. It consisted of thirty-four students engaged in the five AEC global project teams. We identified in our study the re-representation technique to build common ground that is used by team members during multimodal and multimedia communicative events in cross-disciplinary, geographically distributed settings. We define re-representation as a sequence of representations of the same concept using different communication channels and media. This research looks at the role of re-representation as a communication technique in distributed student teams working on AEC projects and how re-representation relates to the timeline of the project and the tasks they try to accomplish. Sharing representations of their concepts and building common ground is a vital part of any project, and re-representation is a vital part of the team process. With a better understanding of how, when, and why re-representation are used, we can better prepare students to address communication challenges and leverage communication channels and media to express their concepts and overcome those difficulties. Our research hypotheses are as follows:

  • Project teams typically capture in meeting minutes the decisions and action to be executed in the next step of the project. Significant sources of information behind decisions and request for actions are embedded within the fabric of communicative events in which participants use both informal and formal media to express their ideas. However, these communicative events that represent the thought process toward decisions are typically not captured in meeting minutes. These sources of information can facilitate common ground building and accelerate the execution of action requests.

  • Re-representations of concepts, i.e., sequences of representations using diverse media and communication channels, mediate and accelerate common ground building.

  • The use of intra- and interdisciplinary re-representations correlates with high team performance, i.e., effective team process and high product quality.

2 Testbed

2.1 Cross-disciplinary, globally distributed teamwork course

The AEC Global Teamwork course is based on the project-based learning (PBL) methodology of teaching and learning that focuses on problem-based, project-organized activities that produce a product for a client. PBL is based on re-engineered processes that bring people from multiple disciplines together. It engages faculty, practitioners, and students from different disciplines, who are geographically distributed. It is a two-quarter course that engages architecture, structural engineering, and construction management students from universities in the USA, Europe, and Asia, i.e., Stanford University, UC Berkeley, Cal Poly San Luis Obispo, Georgia Tech, Kansas University, University of Wisconsin Madison, and CSU Chico, in the USA; University of Puerto Rico, University of Ljubljana in Slovenia, Bauhaus University in Weimar Germany, ETH Zurich and FHA in Switzerland, Strathclyde University in Glasgow, and University of Manchester, Manchester, in UK; KTH in Stockholm, Chalmers University and IT University in Goteborg, Sweden, TU Delft in Netherlands, and University College Cork, Cork, Ireland, in Europe; Stanford Japan Center in Kyoto Japan, Aoyama Gakuin University in Tokyo Japan, and Tsinghua University Beijing, China, in Asia.

The core atom in this learning model is the AEC student team, which consists of an architect, two structural engineers, and two construction managers from the M.Sc. level. This master builder atelier education model creates a situated learning context and builds a number of bridges, between undergraduate and graduate students, between students and industry, between academia and industry, and most importantly across disciplines. An innovative feature of this course is represented by the role the participants play, i.e., undergraduate and graduate students play the roles of apprentice and journeyman, and faculty and industry experts are the master builder mentors. The industry mentors play a key role in providing real-world industry data and feedback to students that increases the authenticity of the PBL learning experience. Each year, there are 4–12 AEC teams in the class. Each team is geographically distributed and has a demanding owner/client that typically wants an exciting, functional, and sustainable building, on budget and on time. The students have four challenges—cross-disciplinary teamwork, use of advanced collaboration technology, time management and team coordination, and multicultural collaboration. The building project represents the core activity in this learning environment. The project is based on a real-world building project that has been scoped to address the academic time frame and pedagogic objectives. The project specifications include the following: (1) building program requirements for a university building of approximately 30,000 sq ft of functional spaces that include faculty and student offices, seminar rooms, small and large classrooms, and an auditorium; (2) a university site where the new building will be build, such as San Francisco, Reno, Madison, Los Angeles, Weimar. The site provides local conditions and challenges for all disciplines, such as local architecture style, climate, and environmental constraints, earthquake, wind and snow loads, flooding zones, access roads, local materials, and labor costs; (3) a budget for the construction of the building; and (4) time for construction and delivery.

AEC teams model, refine, and document the design product and process. The students learn to regroup as the different discipline issues become central problems and impact other disciplines. They use computer tools that support discipline tasks and collaborative work. The project progresses from conceptual design in winter quarter to 3D and 4D CAD models of the building and a final report in spring quarter. The concept development phase deliverables of each team include the following: two distinct integrated AEC concepts, a decision matrix that indicates the pros and cons of the two alternatives and justifies the selection of one of the two concepts to be developed in spring quarter. The project development phase engages students in further iteration and refinement of the chosen alternative, detailing, modeling, simulation, cost–benefit analysis, and life cycle cost investigation. spring quarter culminates with a final AEC Team project presentation of their proposed solution and reflection of their team dynamics evolution.

All AEC teams hold weekly 2-h project review sessions similar to typical building projects in the real world. During these sessions, they present their concepts, explain, clarify, question these concepts, identify and solve problems, negotiate and decide on changes and next steps. Since the concepts, problems, and challenges are defined by the students who work on that specific project, their level of attention and engagement is maximized. Consequently, the students are highly motivated to exchange and acquire as much knowledge as they participate in the cross-disciplinary dialogue. The interaction and the dialogue between team members during project meetings evolved from presentation mode to inquiry, exploration, problem solving, and negotiation. Similar to the real world, the teams have tight deadlines, engage in design reviews, negotiate and decide on modifications. A team’s cross-disciplinary understanding evolves over the life of the project. The international structure of AEC teams adds the real-world collaboration complexity to the learning environment. A key focus is the effective use of ICT resources to support instruction and learning outcomes, which includes space, time, coordination, and cooperation issues. Most importantly, students learn to use and combine diverse communication channels and media to express and share their ideas and solutions. To view AEC student projects, please visit the AEC Project Gallery at http://pbl.stanford.edu/AEC%20projects/projpage.htm.

2.2 Interactive workspace

To support the synchronous and asynchronous communication, collaboration, and coordination activities of the AEC global student teams, faculty, and industry mentors, we developed an ICT Ecosystem. It addresses the needs of the global teams as they are mobile, create digital content, and engage in interactive creation, capture, sharing, and manipulation of 2D, 3D, and 4D CAD models. The current PBL ICT Ecosystem consists of (1) Network Infrastructure includes LAN/WAN, WiFi, and GSM/GPRS; (2) interactive devices that enable the mobile learners to stay connected with their peers, team members, faculty, and mentors, as well as the content they create and share, e.g., smart cell phones with embedded cameras, tablet PC, Web cameras, SmartBoards, and iRoom (Johanson et al. 2002); (3) collaboration applications that support synchronous and asynchronous communication, interaction and feedback, direct manipulation, knowledge capture, sharing, and reuse. These include commercial solutions such as Skype, MS NetMeeting, GoogleCalendar, GoogleDocs, VSee (VSeeLab.com) (Chen 2001, 2003), and PBL-Lab-developed technologies, such as TalkingPaper (Fruchter et al. 2007), RECALL (Fruchter and Yen 2000), and CoMem (Corporate Memory) (Fruchter and Demian 2002).

The interactive collaboration workspace used in the PBL Lab and remote sites to support the global teamwork activities of the AEC global team members during their weekly meetings include the following:

  • RECALLFootnote 1 collaboration technology and knowledge capture (Fruchter and Yen 2000): RECALL builds on Donald Schon’s concept of the reflective practitioner, (Schon 1983), knowledge creation cycle (Nonaka and Takeuchi 1995), and importance of sketching in communicative events (Tversky 1999) and gestures (Tang 1991). It is a drawing application written in Java that captures and indexes the discourse and each individual action on the drawing surface. Users can create free hand sketches, import pictures, or images of CAD models and annotate them during their discourse. The drawing Java application synchronizes with audio/video capture and encoding through a client–server architecture. At the end of the meeting, the drawing and video information is automatically indexed and published on a Web server that allows for distributed and synchronized playback of the drawing session and audio/video from anywhere at anytime. Users are able to interactively navigate through the session by selecting individual drawing elements and jump to the part of interest.

  • MS NetMeeting Videoconference for application sharing with all the remote sites (e.g., RECALL, CAD, images, GoogleDocs, etc.),

  • a SmartBoard for direct manipulation of content and sketching using the RECALL application,

  • a microphone for audio capture that feeds into the SmartBoard computer that runs RECALL.

  • a Webcam that enables the remote students to see the interactive workspace that covers the workspace in PBL Lab at Stanford.

  • additional SmartBoard, laptop, or projector and projection screen for VSee™ technology (VSeelab.com) for parallel video streaming over the IE browser to enable the participants to see all the remote sites (Chen 2001, 2003).

Figure 1 illustrates an example of the interactive collaboration workspace setting used in Atlantic team weekly meetings.

Fig. 1
figure 1

Interactive workspaces used by Atlantic team during their weekly team meetings

Full size image

3 Theoretical points of departure

The theoretical points of departure for this study include communication theory, design theory and methodology, learning theory and distributed cognition, and knowledge management.

Communication is central to teamwork which is a social activity. The objective is to ensure that what was said or proposed by one member has been understood by all the other team members. In communication theory, this is defined as the process of grounding, or common ground building (Clark and Brennan 1991; Clark and Schaefer 1987, 1989). It refers to the development of mutual knowledge, beliefs, and assumptions that is critical for communication between people. To build common ground, the dialogue needs to have both the presentation phase and the acceptance phase (i.e., acknowledgement that what was said was understood). A positive evidence of understanding comes with continued attention from the participants followed by some type of acknowledgement that what was presented was understood. Clark and Brennan studied conversation dynamics. They indicate that there are two factors that determine how common ground develops, i.e., purpose of communication and medium of communication. People try to establish collective purposes in conversations (Griece 1975; Isaacs and Clark 1987). Clark and Brennan offer in their common ground theory an example that is relevant to our study, called grounding references that focuses on objects, for instance, in our case, team members referring to building components as they design a facility. One of the techniques to achieve grounding references is through alternative descriptions, i.e., when a speaker refers to objects, they use one or more referring expressions such as a description, noun, and pronoun. A way by which the conversation partner can indicate that they can identify the object is by presenting an alternative description. Our study builds on these concepts to analyze multimodal and multimedia communicative events in cross-disciplinary, geographically distributed teams. Similar to the alternative descriptions technique found in conversation dynamics analysis (Clark and Brennan 1991), we have identified the re-representation technique to build common ground used by team members in our study of multimodal and multimedia communicative events in cross-disciplinary, geographically distributed teams. In the context of our study, we define re-representation as a sequence of representations of the same concept using different communication channels (e.g., speech, gesture, sketch) and media (annotation, diagram, drawing, 2D CAD, 3D CAD, video) for diverse activities (e.g., clarification, explanation, problem solving, etc.).

The design of the AEC Global Teamwork course and information and collaboration technologies (ICT) build on cognitive and situative learning theory. The cognitive perspective characterizes learning in terms of growth of conceptual understanding and general strategies of thinking and understanding (Dewey 1928). The situative perspective shifts the focus of analysis from individual behavior and cognition to larger systems that include individual agents interacting with each other and with other subsystems in the environment (Greeno 1998; Goldman and Greeno 1998). Situative principles characterize learning in terms of more effective participation in practices of inquiry and discourse that include constructing meanings of concepts and uses of skills. Teamwork, specifically cross-disciplinary learning, is key to the design of the AEC Global Teamwork course. Students engage with team members to determine the role of discipline-specific knowledge in a cross-disciplinary project-centered environment, as well as to exercise newly acquired theoretical knowledge. It is through cross-disciplinary interaction that the team becomes a community of practitioners—the mastery of knowledge and skill requires individuals to move toward full participation in the sociocultural practices of a larger AEC community. The negotiation of language and culture is equally important to the learning process—through participation in a community of AEC practitioners; the students are learning how to create discourse that requires constructing meanings of concepts and uses of skills.

Since the teams are composed of students from different disciplines, each student acts as a professional in the sense that Charles Goodwin uses it (Goodwin 1994). Being professionals in different fields, each student brings a different professional perspective to bear on the design of a building. This is revealed in how they represent and shape the design of their building (Goodwin 1994). We looked at the use of professional vision in the context of a team of interdisciplinary professionals, expanding the focus from professionals talking to impartial audience (e.g., the jury or industry mentors) to how the students used their professional representation language, goals, constraints, and objectives in professional-to-professional communication. The representations that the students used to explain their professional vision were also important. Their sketches, speech, diagrams, and models were important not just for professional vision or reflection; these external representations formed part of each student’s and the whole group’s cognitive activities, as Hutchins showed in the analysis of pilots actions in cockpits and Zhang in the study of students solving the Towers of Hanoi (Hutchins 1995; Zhang 1997).

Schon’s, Goodwin’s, and Hutchins’ theories focus on intra-disciplinary of practitioner activities, i.e., architects, archeologists, or pilots (Goodwin 1994; Hutchins 1995; Schon 1983; Zhang 1997). Our study focused on interdisciplinary project teams. Building common ground and holding a shared representation of the design were critical to achieve a high-quality design and experience a high-performance team process. Theory of restructuration from the education domain suggests that ideas or concepts are separate from their representations. These representations can be restructured to help different audiences grasp the ideas (Blikstein and Wilenksy 2007; Jacobson and Wilenksy 2006). A classic example is the GasLab, where the gas laws are re-represented using agent-based modeling (Wilenksy 1999). As we will further discuss in the paper, we observed the students in the AEC teams regularly re-represent or restructure the concepts they were representing so that they could be understood by the rest of the team.

Other researchers discuss re-representation in analogical matching of designs (Yan et al. 2003) and different types of design representations and the significance of hardware, prototypes, and artifacts in developing and conveying ideas (Brereton 2004; Logan and Radcliffe 2204). Brereton organizes the dimensions of design representations differentiating between internal vs. external, transient vs. durable, self generated vs. ready made, abstract vs. concrete. Logan and Radcliffe focus on the different moves between physical artifacts and abstract representations in the design and development of products. Oxman explores the act of re-representation as part of the architect’s process of concept development and adaptation of past cases (Oxman 1997). Representation and re-representation in the form of multiple representations of the same design objects were explored in the context of design and emergence using collaborative computer-aided design tools. (Damski and Gero 1994; Rosenman and Gero 1996).

This research builds on Donald Schon’s concept of the reflective practitioner paradigm of design (Schon 1983). Schön defines the process of tackling unique design problems as knowing-in-action. To Schön, design is an action-oriented activity. However, when knowing-in-action breaks down, the designer consciously transitions to reflection activities, called reflection-in-action. Schön argues that, whereas action-oriented knowledge is often tacit and difficult to express or convey, what can be captured is reflection-in-action. This concept was expanded into a reflection-in-interaction framework to formalize the process that occurs during collaborative team meetings (Fruchter et al. 2007). Schon studies the interaction between an architect-teacher helping a student with a building design. As the student runs into a dead end with her design and the teacher engages in reflection-in-action, sketching new ideas, reflecting, and sketching new ones, the student watches on intently and reflects on the teacher’s sketches. Our study builds these concepts and analyzes a more complex situation where a group of students who are from different disciplines engage in numerous reflection-in-action and reflection-in-interaction activities.

The reflection-in-action and reflection-in-interaction represent steps in the “knowledge life cycle” that includes “creation, capture, indexing, storing, finding, understanding, and re-using knowledge.” (Fruchter and Demian 2002). Knowledge that is created through dialogue among practitioners, or between mentors and learners represents instances of what Nonaka’s knowledge creation cycle defines through four steps “socialization and externalization of tacit knowledge; and combination and internalization of explicit knowledge” (Nonaka and Takeuchi 1995). The analysis of interaction and learning experiences during project team meetings build on the knowledge life cycle and knowledge creation cycle constructs.

4 Representation and re-representation of concepts used in AEC project teams

Team members express their ideas, concepts, and proposed solutions using their discipline’s representation language, e.g., architects use metaphors, sketches, physical models, 2D or 3D CAD models, engineers use equations, diagrams, sketches, 2D or 3D models, simulations, construction managers use spreadsheets, bar charts, 4D CAD models. They use diverse communication channels such as speech, gesture, sketch, and media, e.g., audio, video, digital images, etc. The choice of communication channel and media determines the level of detail and abstraction necessary to explore or explain a concept. Different representation techniques and media support the exploration of different aspects of the design and direct the focus of the discussion. For instance speech, gestures, sketches, or block diagrams provide an abstraction level that maintains the necessary ambiguity that allows for the exploration of many possible embodiments of the concept and divergent exploratory conversations. 2D or 3D CAD models offer a level of completion that allows for parametric changes and convergent conversations. Insights are gained by the translation from one representation into another representation of the same concept using diverse media and communication channels. Team members use representation sequences mixing communication channels, media, and levels of detail and abstraction to explore and best convey their concepts.

Similar to the alternative descriptions technique defined by grounding theory and used in conversation dynamics analysis (Clark and Brennan 1991), we identified the re-representation technique used by team members in our study of multimodal and multimedia communicative events in cross-disciplinary, geographically distributed teams. We define re-representation as a sequence of representations of the same concept using different communication channels (e.g., speech, gesture, and sketch) and media (diagram, 2D CAD, 3D CAD, images, video). Re-representations drive and determine the breadth and depth of the design process. Re-representation sequences are used in intra-disciplinary or interdisciplinary communicative events. Team members use the following:

  • intra-disciplinary re-representations to express, explore, and better understand and

  • interdisciplinary re-representations to explain and clarify concepts and design impacts across disciplines.

5 Data collection and analysis

We used AEC Global Teamwork course offered in 2008–2009 as the testbed for our study. It consisted of thirty-four students engaged in five AEC global project teams. We performed a longitudinal study of the five teams and collected data of their synchronous and asynchronous communication over a period of four months from mid January to mid May 2009. We present in this paper the data collection and analysis of weekly team meetings that took place online using the interactive workspace setting illustrated in Fig. 1. Weekly project team meetings were 2 h long. Multimedia digital data from these meetings were collected using RECALL, digital pictures, and video taping the meeting from the PBL Lab where some team members were located. RECALL provided a multimedia information source that synchronized the dialogue in the context of the diagrams, drawings, annotations, text, and written notes. Digital images captured important instances of interactions. Video provided an overview of people, dialogue, content, gestures, and interactive technology (e.g., SmartBoard) used during the meetings. We captured approximately 150 h of RECALL weekly team meeting sessions held by the five AEC teams. In addition, the authors made real-time observations during the meetings and collected videos and still photographs. Observations were also made at the beginning of the class when all the students came to Stanford for 5 days; the teams were formed and started to work together in a co-located setting during this team-building kick-off event.

We used video interactive analysis (VIA) (Jordan and Henderson 1995) and RECALL temporal data analysis (Fruchter 2001) to investigate the dialogue, activities, artifacts, gestures, and sketches of the students as they interacted during their weekly project meetings. VIA is a qualitative analysis method that has its roots in the social sciences. It provides a mechanism to define hypotheses and identify patterns in the rich and complex multimedia data that are collected. Video and RECALL capture the social interaction and activities in real time and enable playback and detailed reflection and analysis of the communicative event.

To illustrate the longitudinal analysis, we present examples of data analysis from four team meetings of two teams—Atlantic team and Ridge team. Atlantic team was composed of an architect from University of Puerto Rico, three structural engineers—one at Stanford University and two at Bauhaus University in Germany— two construction managers at Stanford University and at University Wisconsin Madison. Ridge team was composed of an architect from University of Puerto Rico, three structural engineers at Stanford University, and two construction managers at CSU Chico and University Wisconsin Madison.

The three main lenses used for data analysis considered: activity performed by the team members and purpose of the representations being used, media or channel of the representations being used, and whether concepts were being re-represented. We distinguished between episodes where the re-representations were intra- or interdisciplinary. The schema for the activity analysis was defined as follows:

  • Clarifications: provided by a team member who explained a concept in response to a question or prompt from another teammate, the Professor, or client/owner. This type of response was intended to disambiguate, in contrast to an explanation reply that was triggered by an inquiry question.

  • Explanation: when team members were prompted by an inquiry question. Explanations provided the motivation and reasoning behind a proposed idea or solution.

  • Exploration: when the team was engaged in exploring alternatives (e.g., “what are the options for solar energy?”).

  • Problem Solving: when the team is trying to solve a specific problem (i.e., “what should be the floor-to-ceiling height?”). Exploration engages the team in divergent conversation and design thinking vs. problem solving and evaluation that is driven by convergent design thinking.

  • Closed Questions: when one teammate asks a question that requires a yes/no answer, or a specific numeric data or information that is answered quickly, resolving the question (e.g., “How tall is this retaining wall?”).

  • Feedback: when someone gave feedback on what someone else said.

  • Presentation: when one team member talked about a subject that was not prompted by another teammate. This is different from *explanation* in that the person was not explaining the reasoning or motivations for what they were presenting, merely presenting (e.g., “Here is the model. Here is the path and here is where the trees are…”).

  • Negotiation: when the team was discussing the tradeoffs among possible options.

  • Resolution: the result of a negotiation in the process of choosing an option.

  • Other: all the other activities, e.g., technical, scheduling, etc.

For the activity analysis, the RECALL and video footage was broken up into segments based on the activity of the segment. The length of the segments varied from less than a minute to several minutes, with a few cases that were longer than 10 min. Segments were only categorized if the team was engaged in on of the defined project activities, as opposed to technical problems, scheduling, etc. These items were lumped together in the “other” category.

The activity analysis provided us with an evolutionary overview of the transitions and process transformations of the teams as they moved through the three phases of the project—needs identification, concept development, and project development and delivery. Figure 2 illustrates results of the longitudinal activity analysis during weekly meetings of Atlantic team and provides samples from the three project phases. We selected three team meetings: January 29, 2009—from the start of the project when the team identified the needs of the client/owner and the project specifications; February 26, 2009—from the concept development phase when the team was exploring alternatives; and April 19, 2009—from the project development phase when the team was focused on detailing, modeling, simulation, life cycle cost, and cost–benefit analysis.

Fig. 2
figure 2

Longitudinal activity analysis during weekly team meetings—samples from the three project phases: January needs identification, February concept development, and April project development and delivery

Full size image

It is interesting to observe the variation of time spent engaged in different activities in the three project phases, such as, clarification, explanation, and most importantly the significant increase in problem solving from phase one to phase two of the project, and emergence of negotiation and resolution in the third phase of the project.

The second lens focused on communication channel and media. The schema for the communication channel analysis considered the following categories:

  • Speech: team members talking.

  • Annotation: adding new information to an existing diagram. Annotations require the context of the diagram for the annotating to make sense. Annotations include writing and drawing.

  • Diagram: image on the screen that carries information, e.g., 2D CAD of a floor plan.

  • Gesture: movement that carries information, e.g., gestures that were made while drawing with the pen on the SmartBoard an arc over a building while talking about the arc of the sun as it moves with respect to the building that is designed.

  • Highlighting: drawing on a diagram to highlight parts of it, without extending the information on the screen in contrast to annotation.

  • Drawing/Sketch: drawing or sketching on the screen independent of other externally represented content in contrast to annotation, i.e., a drawing could be made on a blank screen, whereas an annotation would not.

  • Text: text that is presented on the screen, e.g., on a PowerPoint slide.

  • Writing: text written on the screen for instance on the SmartBoard.

We integrated the communication channel with the activity analysis schemas. For this integrated analysis, the RECALL and video footage was broken into segments of 10 s. Figure 3 provides an overview of the integrated longitudinal communication channel and activity analysis of Atlantic team meetings during the three project phases. It is interesting to note that the speech communication channel is continuous and accompanies almost all other media representations. This integrated data analysis method highlights the switching of activities and mix of communication channels and media in support of the different activities during the course of the team meetings in the different project phases.

Fig. 3
figure 3

Integrated longitudinal media/communication channel and activity analysis

Full size image

The results were further analyzed to identify re-representation episodes when team members were going through intra- or interdisciplinary representation sequences. The transcript of the discourse and screen-shot images were added to the results spreadsheet of the data analysis. Table 1 provides a more detailed example of the integrated communication channel and activity data analysis showing multiple re-representation cycles during a cross-disciplinary communication episode used to build common ground and mediate the decision-making process in Ridge team. Table 2 illustrates an example that provides further details to the case described in Table 1, in which the transcript of the dialogue between the engineers and architect in Ridge team is shown together with the screen shots from RECALL, indicating the types of representations used during the re-representation cycles.

Table 1 Example of integrated communication channel and activity data analysis showing multiple re-representations during a cross-disciplinary communication episode used to build common ground and decision in a team meeting
Full size table
Table 2 Example of data analysis showing the transcript and observations, RECALL screen shots showing diagrams, drawings, highlights, and annotations as part of a re-representation sequence used to build common ground and decision making during a cross-disciplinary explanation episode between engineers and architect of a global team
Full size table

Following are the re-representations (Re-rep) used by the structural engineer and architect during the explanation discourse to build common ground:

  • Re-rep-1 Eng: this shows a 2D CAD diagram of the elevation of the building. The engineer also marked with red lines diagrammatically possible solutions (e.g., tension cables or trusses).

  • Re-rep-2 Eng: to better explain the concepts and challenges, the structural engineer uses a second diagram showing the forces acting on the different structural components. He explains the structural behavior and potential collapse situation by annotating and highlighting elements on the diagram.

  • Re-rep-3 Eng: the engineer uses a third diagrammatic representation to further explain the problem and the need for architectural symmetry based on same span size of the two cantilevers. He explains his exploration of solutions, annotates, and draws on the diagram to support his rationale.

  • Re-rep-4 Arch: to paraphrase the engineer’s explanation, the architect uses yet another fourth re-representation of the problem and solution. He uses a blank page to draw what he understood using his architectural representation language—drawing rectangles to represent the different volumes of the building that are symmetric and connected through a core element.

This re-representation cycle provides an example of the use of representation sequences aimed at building common ground. According to grounding theory, to build common ground, the dialogue needs to have both the presentation phase and the acceptance phase, i.e., acknowledgement that what was said was understood. The engineer’s Re-rep 1 through Re-rep 3 represent the presentation phase, and the architect’s Re-rep 4 represents the acceptance phase in the process of building common ground. The architect’s acceptance is not only an explicit indication that what the engineer explained was understood but that the architect will follow-up and act to address the design problem.

It is important to note that traditional meeting minutes summarize the design problems and design change requirements. These typically are generic statements such as “the architectural design needs to be revised to address potential structural collapse.” Such abstract design change requirements lead to numerous cycles of requests of information (RFI) and requests for clarification (RFC) between team members, in this case architect and structural engineer. RFIs and RFCs in turn lead to significant project delays of days and weeks, before a decision is made. Often, the architect maintains the original concept, and the structural engineer designs a significantly more expensive structural solution. By building common ground through re-representation during a team meeting, such time delays, cost increases, discipline misalignments and design integration issues can be avoided and resolved in real time during team meetings (Table 2).

We identified many re-representation cycles that supported the diverse team activities, such as, speech–gesture, speech–gesture–draw, speech–draw–gesture, speech–diagram–annotate–gesture, speech–diagram–present–write–draw, speech–diagram–annotate–draw, speech–present–annotate–draw, etc. We observed through the data analysis that re-representation was used to support different activities in the various project phases. In the first and second project phases, re-representation was mostly used during explanation, exploration, problem solving to build common ground. During the third project phase, re-representation was mostly used during justification, clarification, negotiation, and decision making. The integrated data analysis of the activities and communication channels revealed that re-re-presentation plays a key role in common ground building. Rather than simply serving as a final objective of showing their concepts by presenting them through one of the multiple communication channels, re-representation cycles mediate communication and explanation of concepts.

The data analysis indicates a correlation between use of intra- and interdisciplinary re-representation, team process, and performance (TP); building systems integration and final product quality (PQ). Table 3 shows the TP and PQ for each of the five teams and the typical average percentage of intra- and interdisciplinary re-representations during team meetings in the three phases of the project, i.e.,

Table 3 Intra- and interdisciplinary re-representation, team process and performance, and building systems integration and product quality
Full size table

  • needs identification phase—where intra- and interdisciplinary re-representations took about 40% of the discourse time during team meetings in the best case;

  • concept development—where intra- and interdisciplinary re-representations took about 80% of the discourse time during team meetings in the best case; and

  • project development—where intra- and interdisciplinary re-representations took about 70% of the discourse time during team meetings in the best case.

The TP and PQ of each team were evaluated by the instructor and industry mentors.

Teams that used extensively intra- and interdisciplinary re-representations had high TP and PQ, i.e., Central, Atlantic, and Ridge team.

6 Conclusions

A better understanding of how, when, and why representations and re-representation of concepts are used can provide insights into the complexity of the discourse in global teams and the critical relationship between communication process, activities, communication channels, and media. We identified the role of the re-representation technique in building common ground during multimodal and multimedia communicative events in cross-disciplinary, geographically distributed settings. We presented a methodology for longitudinal and integrated data collection and analysis that links activity analysis, communication channel analysis, intra- and interdisciplinary representation, and re-representation analysis.

Our findings show that the regular use of RECALL during the project meetings was not only beneficial for research data collection but also provided a valuable means to build common ground and communicate important issues in context to team members who could not attend a meeting. For instance, the architect of Atlantic team was not able to attend a meeting during which critical an issue was identified related to the design she proposed. The team proposed three design change options. Since RECALL captured the multimodal and multimedia communicative event, the architect was able to playback the session and understand not only what the options of the design change request but also the rationale and common ground the rest of team developed. Based on that, she was able to proceed and revise the design. In a typical project team situation, where such a RECALL knowledge source is not available, a series of requests for clarification cycles are triggered that delay the time to market. We observed many such instances over the last decade since RECALL was deployed in the AEC Global Teamwork course. These scenarios validate the first hypotheses of our study that capture, sharing, and replay of multimedia communicative events can facilitate common ground building and accelerate execution of action requests.

The Ridge team example is an illustration of best practice in building common ground using multimodal and multimedia re-representation cycles to mediate communication and explanation of concepts across disciplines. It is important to note that in the Ridge team example, the common ground-building cycle took 22 min from identifying a cross-disciplinary design issue, engineer’s explanation of the problem through a sequence of re-representations, architect’s acknowledgement of built common ground, to reaching a resolution and plan for action (Table 1). This and many other such re-representation episodes identified in our study support the second hypothesis of this study, i.e., re-representations of concepts, i.e., sequences of representations using diverse media and communication channels mediate and accelerate common ground building. Other teams that did not use the re-representation technique to explain issues and build common ground required 2 or 3 weeks to take action and revise the design, and, in some cases, the design was not revised leading to increased cost and lower quality of the final product.

Last but not least, the observations and results of the five AEC global teams participating in the testbed in 2008–2009 support the third hypothesis, i.e., use of intra- or interdisciplinary re-representations correlates with team performance, i.e., team process and product quality. Ridge, Central, and Atlantic teams who used extensively the re-representation technique to build common ground reached a high-performance team process and produced high-quality design and construction proposals. Ridge team won the Swinerton Sustainability Challenge competition, which is an industry-sponsored challenge for which all teams are evaluated by the industry mentors who judge the teams’ solutions in terms of sustainable design and construction criteria.

The paper presents examples of cross-disciplinary, geographically distributed teams engaged in concept and development phase of a project from the design and construction industry. Future studies may consider scenarios from other vertical markets and collaboration needs that encompass all phases of the life cycle of a project and product including diverse stakeholders, i.e., from financing to operation and maintenance, as well as distribution to numerous multicultural customers.