Jessica J. Marquez

Evaluation of human and automation/robotics integration needs for future human exploration missions

NASA employs Design Reference Missions (DRMs) to define potential architectures for future human exploration missions to deep space, the Moon, and Mars. While DRMs to these destinations share some components, each mission has different needs. This paper focuses on the identified human and automation/robotic integration needs for these future missions. The outcomes of our assessment is a human and automation/robotic (HAR) high-level task list for each of the four DRMs that we reviewed (i.e., Deep Space Sortie, Lunar Visit/Habitation, Deep Space Habitation, and Planetary), as well as a list of common critical HAR factors that drive the design of HAR integration.

Survey and assessment of crew performance evaluation methods applicable to human spacecraft design

Space is an unforgiving environment where the actions of the crew play a critical role in their health and safety. Given the limited number of crewmembers typically onboard a spacecraft and the multitude of complex systems they must operate, the performance of each individual is of paramount importance. Spacecraft habitat layout and operations are two main drivers affecting crew performance efficiency. Having the capability to analyze and compare crew performance across various spacecraft configurations can help identify improvements early in the conceptual design process where changes are less costly to implement, ultimately reducing overall project costs and improving long-term operations of the system. Currently, there are few comprehensive methods readily available for evaluating crew integration within a spacecraft in the conceptual design phase. In order to address this shortcoming, the goal of this work was to analyze various specialized evaluation methods found in analogous industries that have potential application to human spacecraft design. A survey of more than 400 human performance evaluation methods was completed. Over twenty different attributes were identified for each method and a variety of analyses were conducted to characterize and evaluate their potential use for assessing human spacecraft design options. The analysis revealed a particular deficiency of quantitative evaluation methods that are applicable early in the systems engineering design phase. It also identified five existing methods that could be supplemented to achieve the needs of an early design evaluation method. Additional discussion describes potential issues that must be overcome when developing a method specific for use in human spacecraft evaluations.

Increasing crew autonomy for long duration exploration missions: Self-scheduling

Over the last three years, we have been investigating the operational concept of crew self-scheduling as a method of increasing crew autonomy for future exploration missions. Through Playbook, a planning and scheduling software tool, we have incrementally enabled the capability for Earth analog mission crews to modify their schedules at various levels of complexity. Playbook allows the crew to create new activities from scratch, add activities or groups of activities from a Task List, and reschedule or reassign flexible activities. The crew is also able to identify if plan modifications create violations, i.e., plan constraints not being met. This paper summarizes our observations with qualitative evidence from four NASA Extreme Environment Mission Operations (NEEMO) analog missions that supported self-scheduling as a feasible operational concept.

Integrating human performance measures into space operations: Beyond our scheduling capabilities?

Current planning and scheduling software tools for International Space Station (ISS) support different flight controller teams as they plan daily space operations. Planning and scheduling tools capabilities include integrating digitized ISS state inputs, evaluating their expected future states, and propagating them over time. Extensive, custom-made computational models of operations, of objectives, and of operational constraints help ISS flight controllers identify where scheduled events violate constraints. Based on the current capabilities of these tools, this paper proposes how human performance measures could be better integrated into planning and scheduling tools for space mission operations. Future integration of human performance measures could be applied to state inputs (in this case, the astronauts state) and to modeling human performance operational constraints & operational objectives (i.e., assigned activities) with parameters that are relevant to human performance measures. Gaps between the state-of-the-art for human performance modeling and planning tools for future exploration missions are identified.

Minerva: User-centered science operations software capability for future human exploration

In June of 2016, the Biologic Analog Science Associated with Lava Terrains (BASALT) research project conducted its first field deployment, which we call BASALT-1. BASALT-1 consisted of a science-driven field campaign in a volcanic field in Idaho as a simulated human mission to Mars. Scientists and mission operators were provided a suite of ground software tools that we refer to collectively as Minerva to carry out their work. Minerva provides capabilities for traverse planning and route optimization, timeline generation and display, procedure management, execution monitoring, data archiving, visualization, and search. This paper describes the Minerva architecture, constituent components, use cases, and some preliminary findings from the BASALT-1 campaign.

Playbook Data Analysis Tool: Collecting Interaction Data from Extremely Remote Users

Typically, user tests for software tools are conducted in person. At NASA, the users may be located at the bottom of the ocean in a pressurized habitat, above the atmosphere in the International Space Station, or in an isolated capsule on a simulated asteroid mission. The Playbook Data Analysis Tool (P-DAT) is a human-computer interaction (HCI) evaluation tool that the NASA Ames HCI Group has developed to record user interactions with Playbook, the group’s existing planning-and-execution software application. Once the remotely collected user interaction data makes its way back to Earth, researchers can use P-DAT for in-depth analysis. Since a critical component of the Playbook project is to understand how to develop more intuitive software tools for astronauts to plan in space, P-DAT helps guide us in the development of additional easy-to-use features for Playbook, informing the design of future crew autonomy tools.

Measuring Safety and Performance in Human–Automation Systems

defines human–automation systems as those machines that include automation and require human interaction. Researchers have investigated the benefits and unexpected challenges associated with human–automation systems for many years, growing our knowledge and understanding of the different aspects important to appropriately integrate human operators into these systems. However, the implementation of this research knowledge has been limited. In an effort to take advantage of the benefits of automation, it was integrated into existing work flows without fully appreciating how such a shift would change the work itself. Today, we have a much greater awareness of the issues that these systems bring with them. A recent publication acknowledges that the design and evaluation of human–automation systems is multifaceted, including assessment of intricate system design trades with regard to performance (Murphy & Shields, 2012). It is these trades that we must consider when designing new effective, complex aerospace systems. The benefits of automation have been clear and abundant. Even as far back as the 1950s, the superior ability of computers to do certain tasks better than people was acknowledged, such as completing fast and accurate computations (Fitts, 1951). Automation is most indispensable for time-critical tasks when it is humanly impossible for operators to respond quickly enough. For example, within the domain of fault management, automation was required to respond accordingly when the dynamics of a fault were fast (Moray, Inagaki, & Itoh, 2000). Automatic landing sequence allows Mars rovers to safely land on the surface. No human operator on Earth could remotely navigate the spacecraft through Martian reentry simply because of the physical limitations of long-distance communication in deep space. We have reached a point at which complex aerospace systems have critical automation components, inherently making them human– automation systems. As automation becomes more ubiquitous in aerospace, it is important to consider the roles human operators should have within these systems. It is reasoning from a false premise to assume that future aerospace systems will be completely devoid of human interaction. Our safetycritical systems likely will always include human operators at some level, particularly as we transition between existing technologies and more advanced automated systems. We need to enable systems that are safe and perform well, both with and without human operators. As such, we must carefully design our systems to take into account the known and documented disadvantages and limitations of human–automation performance. For instance, skill degradation (Adelman, Cohen, Bresnick, Chinnis, & Laskey, 1993) is just one of many known repercussions of increased automation. With increased use of automation, operators experience reduced performance while completing tasks without automation. Recently, a metaanalysis of several published experiments concluded that, indeed, there is a trade-off between performance, workload, and situation awareness with the inclusion of degrees of automation (Onnasch, Wickens, Li, & Manzey, 2014). The 696300 HFSXXX10.1177/0018720817696300Human FactorsSpecial Issue Commentary