The problem was “real-time” mission planning for a remote Mars ground vehicle, in the face of 12-minute round trip communication delay times imposed by the enormous distance between Mars and Earth (minimum 61 million miles). Current near-Earth missions (on-orbit, Lunar) depend on very tight coupling between ground control and the remote astronauts, but this protocol for planning and execution will clearly not work with the extreme communication delay, especially in such a hostile environment. We were tasked to find an appropriate workflow to overcome the inherent communication delay, and evaluate it by building a system that simulates the communication delays between Earth and Mars.
To learn about how planning could be done on Mars, we researched analogous domains, looking for ones characterized by hostile operations, spatially distributed human teams, and inherent time delays and communications restrictions. Building construction was found as a good surrogate, with a hostile domain that also has significant barriers separating the construction crew from the architects that designed the building. This domain also properly simulates the separation that exists between astronauts executing the plan and those who created it, ground control.
Since surface transit mission planning hasn’t been done since the last Apollo lunar missions in the 70s, we needed some way of understanding the needs and interactions that would be necessary on the Martian surface. We chose to create short storyboards depicting possible interaction concepts, and test them on scientists and Apollo experts to find a direction to begin designing the interfaces.
Early in the design process, we used paper prototypes to simulate the high level interactions and test them on users. We quickly outgrew the medium and a higher level of visual refinement was needed, so we moved on to creating Flash™ prototypes.
To assure that our system was usable and filled both mission planning and mission execution needs, many iterations of the interface were prototyped and tested. In this process, we found that many aspects we thought to be the most challenging initially were easily overcome, while also exposing unforeseen problems.
We implemented fully functioning networked software to simulate and test our system out in the “field”. We created a simulated Martian surface populated with a number of “science targets” to be examined through the development of a route plan and its subsequent execution. An “astronaut” then used a compact handheld computer enhanced with the addition of our planning software, networked on a delay with a system controlled by a “ground control” out of sight on a laptop, to collaboratively explore the field, gaining as many “points” as possible in successfully attaining the science targets. This process gave us a view of how collaborative strategies could be developed in Mars explorations.
