What We are Learning About Communication Protocols
Bill Clancey
NASA/Ames Research Center (University of West Florida)
July 15, 2001
Today begins our fifth day in the hab, and I am ready for my first, interim report. Each evening the Commander asks me for a human factors report, and I find the request both humorous and slightly frustrating. Every day the crater scientists report on stuff they have taken from the field and how it relates to Mars geology and biology. Where is my stuff? Where are my results? It's four days and I haven't written a single scientific paper!
Understanding the relations of people, their activities, tools, and layout of the hab is not like looking for fossils on Haynes Ridge. You do not simply perform an EVA for an hour, and return with rocks that can be sliced and microphotographed for a daily journal. Human life doesn't come in small samples that fit into plastic bags. Studying people, understanding their practices, is much more like viewing a tapestry with complex designs. The incidents of each day are like threads of this tapestry-a little yellow, a strand of brown, a few thicker sections of gray or green. I'm holding a small handful of these strands now, and the Commander asks, "Tell us about the tapestry." The tapestry is huge, woven irregularly, broken in places, sometimes smooth, sometimes wonderfully textured. But the strands and bare threads of each day are usually mundane-we ate breakfast from 735-905, our briefing lasted for an hour-not the stuff of an MSNBC report. Only gradually, over days and weeks, patterns in human behavior will emerge, and then one can begin the even more difficult work of understanding what they mean. Yet sometimes a major event occurs; an incident stops us in our tracks and forces us to realize that something significant is happening. A whole section of the tapestry falls into our hands. Sometimes these gifts of understanding are painful, and that's what happened two evenings ago.
It was a bright and stormy night. The group gathered around the wardroom table at the request of a crewmember, who had just returned from a long private communication with some colleagues. He was concerned about a safety issue in the hab and had sought advice from outside experts. He now confronted the group with the facts and some recommended actions. Each step of his reasoning was bolstered by claims of authoritative ruling he had personally received. The commander had been party to none of this. As the commander attempted to get clarifications and explore options, the concerned crewmember became defensive and claimed that the commander had no expertise in the matters at hand, and that we must follow the advice of the outside experts.
The moment was tense, and very uncomfortable for the rest of us. Why did the crewmember undermine the commander in this way? Was the safety issue as serious as he claimed? What had the experts really said? Was this an interpersonal issue between the crewmember and the commander, a debatable technical matter, or just a miscommunication? We couldn't say, and now the two contenders were vying for our loyalties. "Look at the faces of the people around you," the concerned crewmember said. It was indeed difficult to look at either at them. Look towards the crewmember as he talked, and you appeared to side with him. Watch the commander closely, and you appeared to question his judgment. Two of the crew remained mostly silent, like myself, apparently hoping to avoid being drawn into the hassle. Another person appeared irked, but on the matter of the issue or that we were having such a rancorous discussion at all, was not clear.
Now, one could go further into the style of the argument and individual behavior, but such matters do not concern me now. I believe that the real issue of relevance, what we can learn from the incident, is how communication occurs with the outside world and how decisions are made. Reflecting on the incident the next day, I was reminded of another communications issue that arose during the Haughton-Mars Project in 1999. With these two threads, a generalizable lesson emerges.
During HMP-99, the base commander was in regular communication with outside experts, who advised on how particular instruments should be used to gather data in the crater (a survey of magnetic irregularities). During a briefing to the entire group, the commander provided a mini-lecture about the nature of the "aeromag" survey and the procedures to be followed. At one point, he conveyed some advice he had received from an outside expert. Motioning to one of the crewmembers involved in the survey, he asked, "Did you read John Speck's email?" "No, I haven't read my email yet this morning."
I had captured this interaction on video and showed it in my talks about Mars on Earth. By my interpretation, the incident illustrated that access to to email, and the internet more generally, could significantly change how field science was performed. In particular, the crew on Mars might be in direct contact with their colleagues on Earth, who could help adapt procedures to unexpected situations and interpret preliminary results to redirect exploration. The incident nicely illustrates what cognitive psychologists call "distributed problem solving"-people who aren't physically together jointly working on some task.
One experienced NASA mission controller disagreed with me. By his interpretation, the incident illustrated that all communications should go through mission support, which would then ensure that all members of the crew with a need to know would receive the appropriate information on a timely basis. If mission support had been in control, the crew member who hadn't read his email yet would have received John Speck's advice; the information wouldn't have come up so haphazardly through the commander during a briefing. I countered that the crewmember was getting the advice when he needed it, and a briefing was an entirely appropriate time and venue for sharing information with the crew. The NASA mission controller rebutted that such decisions, about how to conduct the survey, should be made jointly made with the mission support science officer. Thus, having the commander interact directly with John Speck from Mars would take mission support "out of the loop"-so they would lack information and indeed not know that new decisions were being made. I wasn't sure about this, but it seemed unlikely that once more than four or six people were on Mars that Earth could control or even be aware of every investigative decision. At the very least, the incident showed that email and the internet offered options for getting and sharing information that had not been possible during Apollo and hence merited reconsideration of protocols. Indeed, the Arctic Research Station offers significant opportunities for carrying out experiments in alternative communication protocols.
This incident from two years ago, which I had documented and analyzed, and then discussed with colleagues in Houston, was in the back of my mind as I reflected on the argument between the crewmember and the commander. Obviously, something was wrong here-the communication with outside experts had undermined the commander's authority. The crewmember could insist that he now knew the rules, the scientific principles, and insist on the recommendation he was conveying from the experts. What could the commander say in return? He didn't have the same information. And as he continued to ask questions, to try to understand the basis of the advice, the crew member became defensive and attacked the commander's knowledge and ability to make a decision on this important matter.
Thinking about this further, I realize that my interpretation of what happened during HMP-99 was incomplete; I had overgeneralized from the incident. The private communication from John Speck in 1999 worked well because the commander himself was interacting with the expert. The communication two evenings ago did not work well-indeed it precipitated a tense debate-because a crew member was unilaterally bringing what social scientists call "authoritative knowledge" into the group. This put the commander in a difficult situation, to say the least. My new hypothesis is that such communications should not be allowed. Operations advice, whether it comes from mission support or direct email with a colleague, should only be directed to the commander. An outsider should not be telling a crew member, "Here's what I think your Commander should do." In the worse case, a crew member could use outside experts to wrap his own biases in the guise of authority and then galvanize the crew to his side (and we certainly want to avoid situations on Mars missions when a crewmember says, "Do you side with me or the Commander?").
So how was this matter resolved? We radioed the experts and explored the options as a group. The experts' concern was not as severe as the crew member had stated, and indeed their advised plan was close to what the Commander had suggested. The matter was fairly quickly defused, once the experts' opinions were directly heard by the crew. Everyone had the same information, and a consensus was now possible. If we had a science officer, he or she might have served as an intermediary to present and weigh the information. Whether the group should have been as involved in the final decision making is unclear. This incident provides evidence in for open discussion, another thread later this field season or years later might suggest a different weave.
At the same time, I have been conducting a simple experiment with outside experts to explore the boundaries of what is possible and permissible. This involves non-operational, scientific advice. In particular, we are paying the Canadian Arctic Weather Service to provide forecasts specific to Haughton Crater. Furthermore, for a fee, they answer my questions about the weather and the forecasting process. The idea behind this experiment is that a Mars crew is unlikely to be expert in every area of human knowledge relevant to Mars science. In this case, the meteorology of Haughton is part of the scientific study of the crater, yet we have no trained meteorologists in the crew. Further, we are able to provide observations that constitute feedback and additional data of value to the Weather Service. For example, a photograph of cloud formation during changing weather conditions is useful to forecasters. During this process, the experts back on Earth (the Arctic Weather Service) are devoting more attention than usual and hence learning about the conditions on Devon Island. Thus, the crew and weather service are engaging in a distributed collaboration with mutual learning.
So how is a crewmember communicating with the Weather Service different from the other examples I have mentioned (the HMP-99 aeromag email and the case this week of a crewmember advising about safety)? On the face of it, the weather example violates the hypothesis that only the commander should be communicating with outside experts directly. Certainly, it violates the mission controller's assumption that crewmembers would not be in direct contact with outside experts. So what is different, and why does it appear to work so well?
The difference in these examples is that meteorology falls within the realm of scientific study, which is ongoing, and already resides as a responsibility outside of NASA. The same is already true of Mars-some scientists are studying the weather of Mars and making forecasts on a regular basis. Allowing a crewmember to consult with these experts independently, without seeking the commander's permission or involving him or her in every interaction, is not removing the commander from the information loop for several reasons: 1) the interactions concern scientific data and interpretations, not operational decisions (the Weather Service is not involved in planning our EVAs), 2) the interactions are very detailed, specific to natural phenomenon (e.g., why a low pressure system is stalling), and again not directly related scientific operations (e.g., how we should maintain or operate our local weather station), and 3) the crewmember makes transparent his interactions with the outside experts by mentioning periodically what he is learning and how it is valuable to the mission (e.g., improving our independent ability to interpret satellite photos). In short, this interaction has not proved to be problematic and suggests further that crewmembers on Mars be allowed to consult with Earth experts on matters of data interpretation (to be distinguished from operational advice).
These examples illustrate also how the study of living and working on Mars proceeds in an analog setting. First, the evidence builds and is reinterpreted over multiple field seasons. I am often reminded of previous field seasons, from which I can articulate similarities and contrasts in our experience. Second, we learn by carrying out experiments, such as in setting up the communication protocol with the Weather Service. These experiments are structured to fit in important respects a Mars situation (e.g., no meteorologist in the Mars crew). Third, we deliberately explore the space of alternative designs when technology changes, allowing new forms of human interactions. In this case, I conceived the Weather Service experiment as a way of testing the mission controller's claim that no direct communication should be allowed with outside experts. Fourth, our study is opportunistic, looking for incidents that arise in the "natural experiment" of our crew living and working in the hab-precisely what happened a few evenings ago.
Finally, scientists in the field often find phenomena that they haven't studied before, but which other people may have described and analyzed. As patterns emerge in a study, salient issues become important themes meriting literature search and discussion with other experts. For example, I might present this example to space human factor specialists in Houston and seek their comments and advice for exploring the developing communication protocol. As this essay suggests, a simple-minded protocol is unlikely to be necessary or optimal, but defining procedural boundaries and enforcing them may prove problematic in this world of wireless networking and extensive international collaborations.