HMP98 -- Tuesday, July 21, 1998 

(for more information: Haughton-Mars Project Home Page) 

Written by Dr. Bill Clancey of NASA-Ames Research Center...

I came to the Arctic for the adventure, to explore and learn about a new land, which in my opinion is why NASA exists. As the leader of the Human-Centered Computing effort at Ames Research Center, I came to Haughton Crater with two general plans in mind. First, I wanted to study and describe the daily course of expedition life, to establish a baseline from which we might devise new tools for scientific expeditions on Mars. Second, I wanted to document and reflect on the use of mobile computer technology brought to the field by my colleague, Rick Alena.

Computer scientists generally share the public's excitement in having the latest and greatest technology, whether it be a large laptop screen or a wireless communicator (such as a cellular phone). The difference of course, is that computer scientists not only use the tools they build, they are involved in inventing new kinds of hardware and software. This process of invention may, broadly speaking, proceed from two directions: One may start with theoretical possibilities and see what you can build (such as a faster computer or a software program that automates routine work). Or you may start with people, study how they do their work, and understand what tools would help them. Human-centered computing starts with this second approach--we observe people first-hand, talk to them, and invite them to collaborate with us in inventing new computer tools.

This brings us to my first plan in participating in the Haughton expedition. I carried out, to the extent possible in this often windy and cool setting, a study of work practices. This is a "bottom-up" process of participating in daily life and looking for patterns. You describe and think about all the particulars that most people take for granted--how the camp is laid out, how the space is used, how the daily routine is planned, how people are showing each other and talking about rock and geological formations, how results are shared over dinner conversation, and so on. At the same time, I was considering how I might organize these observations in a computer model that would simulate people's interactions in the field (for example, larger groups might cover less territory in a day's journey, but more thoroughly explore a given area).

To make a long story short, I discovered that the modeling ideas we had previously developed in business settings could apply to scientific field expeditions, but they will need to be adapted. In an office setting, work is more routine; forms and customer orders flow in a regular way from person to person and building to building. Scientific work is different; rather than producing the same product every day, scientific work involves continuous learning and customized planning. For example, each day we go out on excursions with our all terrain vehicles (ATVs). We call these trips "traverses." A traverse, unlike most office work, is a kind of activity that is organized as a kind of special event, which is designed on the spot, with customized destinations, participants, duration, equipment, and so on. It is a kind of expedition within an expedition. Describing traverses in a computer model will require extending how we have previously described groups and activities in office workplaces. I was a little surprised to make this discovery, but I am excited about how my observation of "organized activities" in the field has changed my thinking about how everyday work occurs. This new framework will enrich the simulation tool we are developing at Ames and enhance our capability to model how scientists will work on Mars.

Besides studying traverses, I kept alert for other activities in our expedition life. For example, when George Martin (KSC Flight Surgeon) arrived this morning, we all gathered around to introduce ourselves and tell the story of our camp. This activity of "introducing a new member to the group" is an important time for the group to hear again each person's background and role, so we are effectively meeting each other again. Understanding how a group identity forms by such story telling is important for fostering the cooperation and good attitude that will be an important part of extended missions to Mars.

Related to my study of activities, I also looked at how people congregated in camp-where did conversations occur? Where did we plan the day and learn more about each other's discoveries? No one is surprised that the kitchen tent is where most group conversations occur, but who could have anticipated that the ATV staging area would be the center of the camp? We habitually line up the ATVs outside the work tent (where the computers and communications are arranged). People stand around the ATVs when they arrive back from a traverse or are arranging their equipment for a new traverse. The ATVs also have long, soft seats, which provide the best lounging chair in camp. So you often find people having conversations on or about the ATVs. Anthropologists note how space is used in this way, revealing how the arrangement of space influences conversations, which is where learning occurs. Having found these patterns, we can convey them to computer scientists and other camp designers, who might then get new ideas for future camp designs and tools. For example, we see here at Haughton that we need a larger work tent or perhaps a library-sitting room with more shared laptops. Gathering together over laptops would be another place for conversations about biology, geology, meteorites, and Mars. In human-centered computing we call this "design for learning."

Rick Alena using the Mobile Field Station under black cloth to shield from bright sun. Note ball camera mounted on front right rack of ATV.
This brings me to the second aspect of my work at Haughton, namely to study how the mobile technology we brought from Ames was used and could be improved. Rick Alena and I teamed up with Charlie Cockell, a biologist, over the course of three days of traverses to "Arctic oases." These are green areas a few feet across, with moss and Arctic poppies. Charlie came to Haughton to study lemming burrows in these oases. We found that Charlie organized his thinking around the idea of writing a scientific paper (to be co-authored with Pascal Lee). Charlie's ideas about what observations to make and where to look in the crater were shaped by the expected format of a scientific paper. For example, Charlie saw that the soil was quite deep and then began to wonder about how the temperature varied: Why were the lemming burrows always about 9" deep, while the soil was 18" deep? At what depth did the coldness of the permafrost become apparent, making the soil colder? Charlie organized this temperature data into a table, and frequently spoke about figures, sections, and comparisons that would be part of his paper. After the first traverse, Charlie's paper was about the oases and not lemmings as he first thought. Our thinking changed, too. Originally, we conceived of a "field observation record" in terms of text, photographs, and GPS (global location) records. Now, here in the field, we could see that Charlie organized his thoughts in a much more specific way--according to the parts of a scientific paper. Working with Charlie and his colleagues, we will now be able to invent a better software program that fits what a scientist is thinking and doing in the field.

We see in the short examples I have given tremendous progress in less than week of field research. We have applied human-centered computing principles by starting with people. We have learned how they think about their daily work, how they plan and organize themselves in space and time, and how they gather data and form a product that can be shared with the broader scientific community. We will return to Ames to model traverses and other expedition activities, and then attempt to adapt these models in a planning and scheduling tool for Mars missions. We will reconsider the field observation record idea and perhaps add two broader collaboration tools: 1) multimedia methods by which people in different expedition phases can communicate with each other (what we call "corporate memory" in office workplaces), and 2) an internet link by which scientists in the field can access publications from the field and show colleagues back home what they are seeing.

With several days left, it is already apparent that this experience has been a total success. As a computer scientist I have established my first working collaboration with NASA field scientists, and I have enjoyed a great adventure.