Comparing Architectures - Pressurized vs. Unpressurized Rovers

Sensitivity Analysis and Conclusions

Page:  1   2   3

Sensitivity Analysis: Relative impact of 10% performance improvements

Here, we explore the relative impact of a 10 percent improvement in performance of each of the sample-collection activities included in our study, along with driving and deployment of the science station, to determine which performance improvements would yield the greatest increase in productivity. We look at the effect of varying one parameter while keeping all others constant (i.e., calculating a partial derivative) for all parameter variables comprehensively and very rapidly (a computer processes this equation within a very small fraction of a second). The following formula was used to perform this calculation:

As seen in Table VI, two points stand out: First, the order of importance of the drivers is extremely robust. Within each scenario, the order remains consistent throughout the study's variations in time margins and number of science activities. With the sole exception of the one-time activity, "deploy science station," the order of importance of the various drivers also remains consistent under both the 2UPR and 2SPR scenarios. Reducing the time required to deploy the science station has a much greater impact in the 2UPR scenario than in the 2SPR scenario because in the 2UPR case, that activity occupies a much larger percentage of the total work done.

Table VI. Sensitivity Analysis

Activity Agent Impact: 2 UPR Impact: 2 SPR Reference Case Driva EVA Astronauts EVA 100.0 100.0 Driva EVA Astronauts n/a 65.6 Deploy science station Astronauts EVA 4.4 17.3 Trench Astronauts EVA 3.2 32.3 Collect rocks Astronauts EVA 3.2 32.3 Collect regolith Astronauts EVA 3.2 32.3 Use drive tube Astronauts EVA 3.2 32.3 Rake GC teleop 2.2 21.8 Drill soil core GC teleop 1.9 19.4 Drill rock core GC teleop 1.2 11.6 Ref + Margins Driva EVA Astronauts EVA 100.0 100.0 Driva EVA Astronauts n/a 74.1 Deploy science station Astronauts EVA 2.7 21.9 Trench Astronauts EVA 2.0 40.9 Collect rocks Astronauts EVA 2.0 40.9 Collect regolith Astronauts EVA 2.0 40.9 Use drive tube Astronauts EVA 2.0 40.9 Rake GC teleop 1.3 27.5 Drill soil core GC teleop 0.6 24.5 Drill rock core GC teleop 0.4 14.6 Ref + Margins + New activities (blue font) Driva EVA Astronauts EVA 100.0 100.0 Driva EVA Astronauts n/a 68.4 Deploy science station Astronauts EVA 3.1 22.2 Trench Astronauts EVA 2.1 41.3 Collect rocks Astronauts EVA 2.1 41.3 Collect regolith Astronauts EVA 2.1 41.3 Use drive tube Astronauts EVA 2.1 41.3 Rake GC teleop 1.5 27.9 Drill soil core GC teleop 0.7 24.9 Ground-Penetrating Radar GC teleop 0.6 19.8 Drill rock core GC teleop 0.4 14.8 Lidar mapping GC teleop 0.4 14.8 Microscopic imaging GC teleop 0.1 2.5

("GC teleop" means teleoperation of a rover by ground control on Earth. The downward-pointing arrow indicates that "deploy science station" belongs lower in the list for the 2SPR case.)

Second, the greatest impact on productivity by far of a 10 percent improvement in any single activity is seen in driving. Decreasing the time required to drive any given distance would mean that additional time could be spent in productive work at the science-activity sites. Note, however, that a different geometry with smaller distances between localities would result in driving speed having a lower relative impact, while longer distances would exaggerate the importance of driving even beyond that of this study.

The next-highest impact level is seen in 4 activities that are conducted by astronauts working in EVA mode. Since EVA time is expensive, a 10 percent reduction yields a substantial increase in productivity. A smaller, but still significant level of impact appears in the 3 tasks performed by the SPRs. Improving the performance of activities conducted by the UPRs produces relatively little improvement in productivity; the time saved in carrying out these tasks is overwhelmed by the time spent in commuting to and from the lander-habitat.

If desired, a higher degree of resolution can be achieved by decomposing each function into its component functions. For example, the activity "drilling a soil core" consists of selecting a site to drill a sample (instrument technology), mobility (getting to the sample), manipulation (acquiring the sample and packaging it), etc.

Conclusions

• An automated planner is very useful in managing a large trade space and finding non-intuitive solutions.

• SPRs are significantly more productive than UPRs, given the study's constraints and assumptions.

• The 2UPR scenario could be more productive if more activity sites were close to the pressurized lander-habitat.

• Investing in improvement of driving performance produces a much greater productivity increase than improvement of any other activity.

• In the 2SPR scenario, our expanded list of activities still does not fill the available time in the 14-day mission.

• Sample mass in the 2SPR scenario is likely to exceed the capacity of the return flight and so require triage.