This summer we focused on improving communication between scientists and engineers to assist with mission planning during a simulated robotic rover mission. The team created a collaborative software prototype designed to help scientists better accomplish their science objectives. By using visual planning tools that link intent with instrument settings, this software allows scientists to more easily convey their scientific goals to the flight team.
The following is a discussion of the success of our prototype in meeting this goal during our final Landing Day simulation (July 24, 2009), and recommendations for future work based on our research and observations throughout the spring and summer.
Our final design involves a major change in workflow. We propose having a geologist actually create the plan so they can accurately represent their plan intent, instead of dictating what they want to an engineer who lays out the plan for them. In order to allow scientists to make their own plan, we strived to make our tool very straightforward and approchable. During the “Landing Day” scenario, users found the tool intuitive and easy to use. One user commented that many tools get more complex with more features, but he found our tool to be very simple straightforward. We found that all users with whom we tested the tool, whether they had planning experience or not, were able to start laying out a plan right away.
In order to help scientists create a plan themselves, we also provided a visual way to understand instrument settings and easily indicate what sort of observations the scientist wants. Conveyance enables scientists to lay down waypoints directly onto a satellite image and indicate roughly what scale the scientist can expect to see in the image.
Everyone who participated in the Landing Day scenario responded very positively to the field of view interface. Scientists understood intuitively how to interact with the tool and were able to easily plan within the constraints of the rover. In addition, scientist were pleased that they could preview the millemeters per pixel at a given distance. All of our partcipiants, even those from a non-scientific background, understood that the gradient was an indication of what quality image they would get at a given distance.
The main purpose of our tool was to clearly capture the intent of the scientists' plan so that it can be interpreted by the flight team. This is done both visually on the map and in the task list's “notes” field. The notes field successfully helped scientists communicate their planning goals. There were only a couple of instances where science was not happy with flight’s changes. However, these were mainly instances where there was no way flight could adjust the plan to account for both the science goals and the constraints. In this case a voice loop would have been appropriate to discuss contingency plans, but it was necessary to restrict communication for the purpose of our evaluation. Even though many users said they would have liked to have some sort of verbal communication, they were still able to get across the goal of each observation and flight was able to interpret their intent. Users liked the simplicity of the note field. A NASA planetary geologist who participated in our final evaluation commented that the MER mission made an attempt to incorporate notes, but scientists stopped using them because the could not figure out what information was supposed to go where. The ability to simply write out what the goals are, and have it stored directly in the plan proved very successful.
Our prototype only addresses one of the many possible directions we considered focusing on this summer. The following is a list of recommendations for future work based on our research and observations over the past eight months.