An Operational Assessment of Team Cognition for Long-Distance Space Missions

Abstract

Long-duration and long-distance space missions create a complicated context for collaboration. The combination of environmental and psychosocial factors will negatively impact individual and team cognition, potentially jeopardizing mission success. To provide a richer understanding of the issues surrounding team cognition for future space exploration missions, this paper describes an operational assessment to identify the key team cognition factors most likely to affect problem solving. Semi-structured interviews were conducted to elicit team cognitive and associated factors relevant to long-duration spaceflight missions. Ten NASA employees (e.g., astronauts, mission controllers) were interviewed to provide a more complete and accurate representation of individual and team cognitive demands. Our goal was to explicate critical cognitive processes associated with team effectiveness in space missions and provide guidance on technology and training collaborative problem solving for space missions.

Introduction

The context of long-duration and long-distance space missions makes effective collaboration within and between teams extremely complicated. The combination of environmental factors, such as isolation and confinement, coupled with psychosocial factors, such as risk and stress, will negatively impact individual and team cognition, subsequently affecting team effectiveness and jeopardize mission success (e.g., Casler & Cook, 1999; Faerman et al., 2023). Further, astronauts in future space exploration missions will have to deal with extended periods of isolation and confinement as well as communication constraints caused by extreme distances. Accordingly, these teams will need to operate under a much greater level of autonomy than current spaceflight crews. Given that team cognition has been shown to be a significant predictor of team performance across a number of domains and tasks (e.g., DeChurch & Mesmer-Magnus, 2010), it is critical to understand how team cognition occurs under these specific conditions, and what countermeasures, like team training (Delise et al., 2010) might be developed to improve team-level cognitive processes as well as cognition between teams.

We focus on complex cognitive processes engaged by teams, such as planning, decision making, and collaborative problem solving (Fiore et al., 2023). Although well studied by the human factors community (e.g., Jones, 2010; John Paul et al., 2010), research into these team processes under conditions of isolation, confinement, danger, high autonomy, and long durations and/or distances is still limited. As such, we aim to contribute to the growing body of research on team cognition for long-distance space missions (LDSM). Note that although much of what we discuss pertains to both long-duration and long-distance space missions, our emphasis is on long-distance missions. We focus here on the increased need for crew autonomy arising from factors like great distances, isolation, and communication delays, and we are interested in how these will affect collaborative problem solving.

Team Cognition

Team cognition as an area of inquiry is foundational to our understanding of complex cognitive systems. Teams are increasingly composed of humans and machines and are utilized to ease the workload and maximize the benefits of distributed expertise in complex cognitive operations (Fiore & Wiltshire, 2016; Morrow & Fiore, 2012; Salas & Fiore, 2004; Salas, Fiore, & Letsky, 2012). This is particularly the case in human spaceflight. The study of team cognition is advantageous for understanding spaceflight given that with a team comes a suite of collective cognitive resources including knowledge, skills, and abilities, all of which are necessary for mission success in high-stakes situations (e.g., Fiore et al., 2014; Sikorski et al., 2012).

In this paper, we address team cognition with an emphasis on collaborative problem solving (CPS). Our focus on team cognition, rather than individual cognition, is essential for understanding and improving CPS due to the advantages that teams provide in the retrieval of information when communication efficiently provides a form of knowledge coordination. For example, teams allow for “multiple access points or triggers for memory retrieval. . . [allowing team members to] recognize errors in another person’s recollection and [thus] may collectively be able to construct a more accurate aggregate account” (Morrow & Fiore, 2012, p. 205).

Relevant to our focus on collaborative problem solving, is the distinction between behavioral coordination and knowledge coordination. Behavioral coordination is associated with the timing and sequencing of actions in service of team goals and tasks. This typically manifests in time-stressed situations when a team’s procedure needs to be executed safely and effectively. By contrast, knowledge coordination more generally describes the awareness and use of team member expertise. That is, in situations where team member knowledge is distributed, for a given team to effectively execute its tasks, it needs to know where to find critical information and knowledge as well as how and when to apply that knowledge. As such, we consider knowledge coordination and behavioral coordination as key processes to support collaborative problem solving and performance in complex collaborative environments.

As the theoretical foundation for CPS, we draw from macrocognition theory, a term to describe cognition in naturalistic environments, taking both the individual and the environmental factors they interact with, such as equipment or machinery, into consideration (Klein et al., 2003). Our emphasis is on collaborative problem-solving in complex environments. In this case, solving problems involves integrating knowledge across many interconnected factors distributed across socio-technological systems (Fischer et al., 2011); that is, the problem factors exist in the technology, the environment, the team members, and the different teams associated with human spaceflight (i.e., the astronaut team and the multiple teams on Earth coordinating the mission). Many of these problems are complex given that the tasks in which these arise are dynamic, varying as a function of temporal demand, and many of the variables involved do not display a one-to-one relationship (Quesada et al., 2005). Thus, complex problems, by necessity, require teams to collaboratively solve them. As studied in space crews (Orasanu, 2005), this requires the teams to be able to fluidly adapt and develop a robust and well-organized knowledge repertoire, taking into consideration shared cognitive processes within as well as between teams.

Research Purpose

In an effort to support the need for further understanding team cognition for future space exploration missions, this paper describes an operational assessment to identify the key team cognition factors that are most likely to affect the maintenance of effective and adaptive team performance (Burke et al., 2006; Cooke et al., 2007) and overall crew well-being. From this, we hope to extend our understanding of team cognition in complex environments, as well as provide guidance on training needs for long-distance space missions.

Methods

Ten NASA employees from a variety of backgrounds, including astronauts, mission controllers, operations researchers, and training managers, participated in our Operational Assessment. Based upon the team performance literature, with an emphasis on collaborative problem solving (Fiore et al., 2010; Graesser et al., 2018), an interview protocol was developed to elicit team cognitive and associated factors relevant to long-duration spaceflight missions (e.g., Jiang et al., 2023; Kanas et al., 2007). Participants were told that we were working to understand how astronauts and mission control personnel collaborate to solve problems, plan, make decisions, build knowledge, communicate, and coordinate (e.g., Marques-Quinteriro et al., 2013; Palinkas & Suedfeld, 2021). Interviewees were asked to recall personal mission experiences and specific instances that can be used to provide detail. They were also asked to provide examples to help us compare nominal/routine situations with off-nominal/non-routine situations. Further, probes were used to help identify tasks/issues, helping us focus on areas that may be more difficult for mission control to give up control over and where training might be improved upon as we move toward missions requiring more crew autonomy.

Interview Results and Interpretations

We categorized interview segments by the type of cognitive knowledge and processes described. Given space limitations, we are not reporting the full outcome and interpretation of the interviews. Rather, we emphasize key elements with a particular focus on communication and shared cognition in the context of collaborative problem solving, connecting them to extant findings on spaceflight teamwork. Next, we discuss some of the consistent themes cutting across various aspects of mission needs. Finally, we conclude with a discussion of potential technological scaffolds that can support collaborative problem solving in LDSM.

Although much was gleaned about cognition and collaboration, in this brief paper, we focus on communication challenges that will occur in long-distance missions. For example, interviewees discussed a range of processes subsuming cognition and collaboration, including information sharing and synthesis, solution evaluation, as well as decision making and problem solving. A consistent theme from the interviews was that the communication structures currently employed for the ISS-missions paradigm will not suffice for LDSM. The reason for this stems from both the time delay and its disruptive nature (e.g., Fischer et al., 2013; Hagemann et al., 2023), but also the fact that the communications bandwidth between LDSM vehicle and stations will not be the same as they can be on low Earth orbit (LEO) or the Moon. For example, one interviewee described how data storage transfer would be a significant challenge for LDSM, stating that “the data pipe going out to the long-distance crews is critical. Going to Mars with the Curiosity rover, which is receiving tons of data that it can’t send back to Earth because the data pipe is too small; hopefully it will be better when we get humans on board.” Thus, mission control will not have rich data that they currently access from manned spaceflight missions, thus attenuating their awareness of problem elements and limit developing a comprehensive model of the problem space.

These communication challenges will affect the coordination of information and knowledge between ground and crew in problem solving incidents. An interviewee emphasized the importance of this for team effectiveness: “Problem solving is constant and has to be a natural thing that everyone does and the team is good at. . . need to know when to incorporate ground and when not to. . . need to be autonomous in solving own problems and better at expressing what they are seeing, the ramifications, priorities/urgencies, etc. to the ground. . .” Another highlighted more of the extended cognitive system with the role of Mission Control and Flight Director in support of team planning and decision coordination: “The flight director asks what each entity is doing during the day and decides on their plan based on these conversations.”

One interviewee described how he better understood the challenges that would arise from comm delays when dealing with them in analog missions. “We simulated the delay throughout the whole mission, so in the beginning, you could talk live via an intercom, after the first month only text or video communications were available with the communication delay. . . . If you were under communication delays, working on payloads, if you don’t have communications you will tend to not ask the ground control (GC) for small details, you’ll just try to figure them out yourself because communicating with a time delay of 10 min would be less efficient. So fewer communications were going on with the ground.” Similarly, the interviewee noted that GC can limit communications, producing unintended consequences when there is a potential for mistakes. “Why didn’t they tell me this? Maybe I’m doing other things wrong. So that builds stress and makes things difficult.” This becomes particularly problematic when dealing with unique and unanticipated problems.

More proximal to problem solving, one of the flight operations personnel described the heavy reliance on diagrams etc. for the various technical systems. He stated: “There is documentation and flow charts, logic, etc. the teams with those backgrounds are creating. . . Malfunctions are also flow charted.” Additionally, they noted that troubleshooting flowcharts sometimes include an entry related to human input: “[There are definitely] steps the humans would do in the flow charts. If something went wrong, they see [the] signatures so they go here; they literally work their way through them.” These flow charts can be considered cognitive artifacts that entail concrete visual and semantic knowledge and serve as externalized cognition for the problem solvers (Fiore & Wiltshire, 2016). “They’re called malfunction books, they have different names, but some are pretty thick. [For something like a] major power malfunction, you should see the amazing set of steps you should have to go through, that are thought out.” An additional factor, though, is how these are not at all automated and still need to be manually distilled for troubleshooting: “but you need to spell out in paper or PDF format that the flight control or crew would execute. . . [because ground minimizes] what the crew would have to do.” When extrapolating to future LDSM, an interviewee noted how this increasing autonomy driven by distance will affect their problem solving: “The flight controllers are comfortable working technical issues (they are engineers) . . . not sure how a flight control team that is 40 min delay is going to work with a team that faces a problem that was not anticipated before launch.”

These issues align with research on communication delay and coordination during space missions. In a study investigating the effect of 50 and 300 s delays on astronaut and flight controller cooperation, Fischer et al. (2013) found that transmission delays disrupted the structure and efficiency of space-ground communications. Individuals’ contributions overlapped as they did not mark the end of their turns, more so in the case of longer transmission delays. Moreover, inadequate listener feedback and grounding strategies were observed. Another study demonstrated how transmission delays increased the time it took to repair system failures on simulated spacecraft problem-solving incidents (Fischer & Mosier, 2014). Importantly, in the case of time delay, no differences were found in conditions using voice versus text-based communication.

Related to communication outcomes, we found that interviewees emphasized the importance of developing team knowledge structures: “Shared mental models among ground, crew, and family is very important, including schedule and expectations of the mission docking of spacecraft.” This is similar to experimental work where successful teams were shown to establish shared task understanding, indicating that shared mental models regarding the problem-space are crucial for successful problem-solving in asynchronous communication conditions (Fischer & Mosier, 2014). Our interviewees, though, went further in describing the importance of what would be considered transactive memory systems, the collective storing and recalling of information (Palazzolo, 2005). This was highlighted when describing cross-team knowledge coordination: “As the CAPCOM, you are the interface with the crew doing the verbal communication with them. That role is partly communicator and telling them what the team wants to tell them and partly translator because the CAPCOM understands what is happening by crew perspective better than anyone else in the room.”

There were also a number of recurrent themes in the interviews relating to team performance. One theme was the importance of managing social and emotional relationships. This was not only with the flight but also with the ground crew and sometimes even with the flight crew’s families (e.g., Palinkas, 2007). Although more of a distal issue, this can have downstream consequences if limiting openness in communication. More proximally related to CPS was a concern about the amount of technical knowledge crew members will require with increasing independence from the ground during LDSMs. This included the design of the hardware/vehicle to minimize this issue. Some interviewees suggested this would require a vehicle with a design relatively simple enough for the crew to maintain and repair provided they are given the appropriate technical knowledge.

Interviewees also noted the importance of allowing crews to train together and work together for long durations before the mission (cf., Watts-Perotti & Woods, 2007). Although this is customary for NASA missions, it will be increasingly difficult considering the highly probable multinational nature of LDSM. The training would enhance not simply behavioral coordination but also the kind of knowledge coordination needed to support CPS. Lastly, we also saw an emphasis on the need to select crew members who are emotionally independent and able to maintain a high degree of self-motivation over long periods of time. Although another distal factor in the context of CPS, such self-regulatory processes can attenuate the tenacity needed for more challenging problems experienced in LDSM.

In short, our operational assessment allowed us to better understand how critical team cognitive knowledge and processes occur in mission operations and some important issues salient to operations personnel. We now discuss a potential method for scaffolding some of the above communication and collaboration processes. Our goal is to illustrate how an evolution of existing capabilities can be achieved by theoretically grounding CPS in the context of an extended cognitive system.

Cognition and Communication Scaffolds

As one of the most predictive elements of team effectiveness, communication processes and technologies are an important target for interventions and mitigation strategies supporting LDSM. For example, in support of mission success, research has examined how a structured communication template was leveraged to enhance communication protocols (Fischer & Mosier, 2015; Mosier & Fischer, 2023). These protocols were rated as helpful in reducing the negative impacts of communication delays. Nonetheless, while different communication strategies have been developed to ensure more efficient and less frustrating communication methods in the conditions of communication lag, the communication structure itself doesn’t help the astronauts with decision-making processes in problem-solving incidents. Specifically, while GC is currently the main monitoring entity for problems and issues, and the ones who flies the station and makes the decisions, this will not be possible in LDSM. Therefore, the crew must make many decisions on their own to ensure operative problem-solving processes and the safety of the vehicle and the crew.

At the same time, it is important to not forget that GC can still be delegated some problem-solving. Therefore, this requires improved processes and technology to help address problem solving in LDSM. First, these need to establish what kind of information is possible to share with the ground. Second, these need to establish what kind of problems can be solved by the crew themselves or need to be put on hold and allow for solutions from GC. Third, the problem solving process, and the problem solvers need to be made explicit in any technological scaffold. We propose two main sets of guidelines to encourage more effective CPS in such a context: aspects related to diagnosing and categorizing, and processes of coordinating cognition and communication.

Diagnosing and Categorizing

As the first element, we propose a triage-type system supporting crew capabilities in decision making when addressing problem solving needs. Given space constraints, we provide a broad outline for design consideration. Specifically, the system would first need to distinguish between three different types of events: (a) emergencies (red), (b) events requiring a somewhat quick response but not immediate (yellow), and (c) events that do not require a quick response and can be delayed, possibly with enough time for GC input (green).

Given the examples provided in the interviews, and in our review of prior NASA problem solving (e.g., Fiore et al., 2014; Käosaar, et al., 2024), not all problems will neatly fall into one of these categories. As such, the system would need to be able to label and monitor composite events. With composite events, for example, the crew would need to make an immediate decision on their own (red) for some problem elements (e.g., how to engage a partial equipment shutdown), but after the immediate response, the event could be categorized as yellow or green (e.g., identification of repair options).

Coordinating Cognition and Communication

An important complement to diagnosing and categorizing problems is an approach for clearly coordinating cognition and communication. Recent research has studied a communications system for asynchronous communications called braiding (for more detailed description, see Fischer et al., 2023). Braiding is a text-based communication software which allows for continuous conversation between the space crew and mission support personnel during transmission delays. It essentially uses a rotating carousel of “braids” (panels or topics) with each braid assigned to a distinct topic. Remote team members communicate with each other by taking turns adding content to a braid as it moves into view. As such, braiding provides an innovative way to organize communications to help coordinate more general information and knowledge sharing.

We build on this to suggest something similar as a tool to scaffold complex collaborative problem solving. Braiding provides an important grounding for problems requiring awareness of team and task interdependencies. As currently conceived, braiding affords a capability for creating thematically related topic elements. This naturally lends itself to problem solving as problem elements can be parsed and themed depending on their nature. We propose a capability for braiding that adds another dimension concerning team roles and relevant expertise. This overlays a capability supporting shared cognition. Specifically, it scaffolds the kind of transactive memory system identified as important in our interviews; that is, knowledge of who knows what, and/or has expertise in relevant areas. Although such knowledge might already reside in crew members’ internal memories, adding this as a tagging method to braiding externalizes the cognition associated with the transactive memory (cf. Fiore & Wiltshire, 2016). Further, given the turnover associated with mission duration and ground control personnel, transactive memory systems will likely be less detailed, if not entirely lacking. As such, this additional dimension creates a capability for scaffolding shared cognition within and across the teams.

What would next be required is a more direct linkage to deeper, extensive knowledge associated with a problem. As described earlier, interviewees discussed their use of “malfunction books.” As such, these are necessary resources given the complicated nature of space systems technology. But these are not yet neatly integrated with current problem solving, serving more as an external memory aid. Further, they are not specifically associated with problem solvers and serve rather as generic knowledge. Given this, we see these as an additional entry point for machine aiding, where, when digitized, they can be incorporated with the braiding system described above. Specifically, just as braiding would benefit from team-related role knowledge, “malfunction books” could provide much-needed task-relevant knowledge. In this way, braids can be linked to more detailed technical information, affording rapid access to specific problem elements. Additionally, team members could be linked to a braid connected with a problem element and its associated malfunction book. In combination, task-related knowledge is then connected with team-related knowledge inherent in the braids.

In short, we have outlined a way to scaffold cognition and collaboration through a type of human-machine extended cognitive system where knowledge of the problem environment and the team is distributed and demarcated across the people and the technology. In time-stressed situations, rapid access to such knowledge will enable more efficient problem solving due to decreased communication errors and subsequently improve behavioral coordination.

Conclusions

In this paper, we have described expert insights into team cognition related to collaboration in LDSM. We additionally provided a broad overview of communication technology that can support knowledge coordination. This adds to our prior theorizing and research with the “macrocognition in teams model” (MITM). According to the MITM, in service of team problem solving, team members must engage in knowledge building whereby individual team members internalized knowledge is transformed into externalized knowledge by both individual and team-level cognitive processes (Fiore et al., 2010). In this way, teams collaboratively build knowledge, drawing upon their expertise, by transforming data into information into knowledge. The technology we propose would support the cognitive processes of both the individuals within teams, the emergent cognition of the team itself, as well as shared cognition amongst the multiple teams of LDSM. That is, it scaffolds what the MITM views as the parallel, interdependent, and iterative nature of collaborative cognition whereby the formation of knowledge within individuals impacts the formation of knowledge within the team and vice versa (Fiore et al., 2008). As such, research must explore how such technological scaffolds can contribute to effective team problem solving outcomes for LDSM.