Human Performance Metrics for Spacesuit Evaluation

The goal of this project is to address the technological research gaps within Technology Area Breakdown Structure (TABS) 6.2.1 Extravehicular Activity Systems: Pressure Garment and 6.3.4 Human Health and Performance: Human Factors outlined in the 2015 NASA Technology Roadmaps (TRR). This work aims to use biomechanics, bioenergetics, and human-systems engineering to create quantitative human performance metrics. These metrics will also inform the specific research gaps addressed in Human Research Roadmap (HRR). These include: EVA Gap 7 that asks what EVA suit design parameters affect human health and performance, newly proposed EVA Gap 7B that asks how suit sizing affect crew health, performance, and injury risk, EVA Gap 10 that asks what physiological parameters during EVA operations can inform about crew health and performance, and EVA Gap 11 that asks how EVA operations increase crewmember risk of injury. To create these metrics, this work will use portable and wearable devices to effectively compare, quantify, and evaluate human motion and performance in both unsuited and suited conditions during functionally relevant and naturalistic environments. We aim to use the performance metrics gathered from these devices to address the technological and human research roadmaps listed above, specifically regarding quantifying spacesuit fit and performance (TRR 6.3.4.9, 6.2.1.7 and HRR EVA Gaps 7, 7B, 10, 11). To this end, the specific aims of this study are as follows: Aim 1: Use of cognitive task analysis to determine a series of operationally relevant assessments, tasks, needs for decision makers during the evaluation of human spacesuit performance in the context of defining spacesuit design requirements for minimizing risk to crew health. Aim 2: Creation of appropriate new performance metrics and data acquisition modalities to quantify suit fit during operationally relevant tasks. Aim 3: Validation of operationally relevant, quantitative performance metrics with operational decision-makers. While being under this fellowship, significant progress has been made towards narrowing the focus of this work and identifying specific gaps in our knowledge of the human-spacesuit interaction relevant to TRR 6.3.4.9, 6.2.1.7 and HRR EVA Gaps 7, 7B, 10, 11. Aim 1: To accomplish Aims 2 and 3 listed above, this work used a modified Cognitive Task Analysis (CTA) framework. CTA aims to understand knowledge, thinking processes, and goals within observable environments. Utilization of CTA defines the needs of both users and decision-makers for given information processes in operationally relevant settings. For this work, we define the user as the human wearing the spacesuit and the decision-maker as anyone involved in the design or evaluation process of the spacesuit (i.e. suit engineer, EVA scientists, etc.). Inclusion of these needs in the design process maximizes usability, understanding, and implementation of novel evaluation methods, expediting the design iterations of future systems and enhancing the understanding of the human-spacesuit interaction and performance. This fellowship utilized a modified CTA. Through observation of suit fit and familiarization sessions at NASA Johnson Space Center and through discussions with suit design engineers and EVA operation scientists, we have begun to better define the situation awareness (SA) requirements relevant to subjectively finding an ideal suit fit. Currently, suit fit checks and familiarization runs are performed with the Mark III and Z-2 planetary space suits for new astronaut candidates and test subjects at the NASA Johnson Space Center by the EC5 Space Suit and Crew Survival Systems group. Observations of these sessions have helped define how subjects are currently being fit into these planetary suits. Currently, suit engineers observe new test subjects as they perform a series of operationally relevant tasks using subjective feedback from the subject and visual observation to determine if sizing elements needs to be changed. Discussions with the leading EVA scientists at NASA Johnson Space Center have also aided in defining the needs of decision-makers in this realm of EVA performance. Combining all observations, we have begun to develop relevant event workflow diagrams that identify the appropriate areas relevant to suit fit that require better understanding and quantification. This human-centered, end-user approach will maximize the end usability and adaptability of these metrics. Thus far, two main categories of suit fit have been identified: static and dynamic. Static fit refers to how the crewmembers sit and index inside the suit (i.e. how much space there is between sternum and hard upper torso (HUT) or how does the elbow joint of the human align with the elbow of the suit). Dynamic fit refers to how well the human and suit move synchronously or asynchronously. Future observations and discussions with EVA scientists and engineers will further develop these event workflows and suit fit definitions. In addition, this CTA framework assesses the appropriate measurement modalities for candidate metric generation. Quantifying both static and dynamic spacesuit fit will require a deep understanding of the human-spacesuit interaction; therefore, measurement modalities should quantify motion of both the spacesuit itself and the human inside. To appropriately quantify human kinematics within the suit, low profile devices are necessary to avoid inadvertently changing how the human indexes inside the suit. This work has begun exploring wearable sensor networks in the form of inertial measurement units (IMUs), strain gauges, and pressure sensors to collect the data necessary to quantify suit fit in a way that meets the needs of users and decision-makers. Preliminary findings from this CTA were presented at the 46th Annual International Conference on Environmental Systems in Charlestown, SC. Aim 2: There is a need to quantify both static and dynamic suit fit. This section provides the physics-based intuition behind the selection of new metrics for suit fit. Thus far, this work has defined two new metrics to beginning quantifying dynamic fit: 1) the difference in suit and human knee angle (ΔROM) and 2) relative motion between the suit and the knee (ρ_HS^n (t)). The difference in the suit and human knee angles is computed by comparing knee range of motion (ROM) during each step. ΔROM is defined by: ΔROM=ROM_S-ROM_H (1) where ROMS is the ROM of the suit knee and ROMH is the range of motion of the human knee. Positive values of ΔROM are indicative of steps during which the suit had a greater ROM, while negative values of ΔROM represent steps when the human had a greater ROM. ΔROM is computed for each step. Relative motion between the suit and human was quantified by adapting the methodology from rehabilitation medicine that quantified coordination, termed the relative coordination metric (RCM): ρ_HS^n (t)=2 tan^(-1)⁡((Ω_H (t))/(Ω_S (t) ))-90° (2) where ρ_HS^n (t) represents the RCM between the human and suit body segment n at time t. Ω_H (t) and Ω_S (t) are the L2-norm of the angular velocity of the human and suit, respectively. By definition, ρ_HS^n (t) ranges between 90 and -90deg, where ρ_HS^n (t)=0° represents motion in which both the human and suit are moving completely synchronously, ρ_HS^n (t)= +90° represents a movement in which the human is moving while the suit is not, ρ_HS^n (t)= -90° represents a movement in which the suit is moving while the human is not, and values in between represent motions with varying degrees of coordination between the human and suit. The time-series nature of ρ_HS^n (t) allows for the observation of how relative motion between the human and suit evolves over time. Aim 3: To begin validating this new metric for dynamic suit fit, a pilot study was conducted in March 2017 to evaluate lower body kinematics and locomotion at NASA JSC. Participants wore a technical cooling garment (TCG) below the liquid cooling garment (LCG). Five strap-on IMUs (Opal IMU, APDM, Inc. Portland, OR, USA), with embedded accelerometers, gyroscopes, and magnetometers (sampling rate of 128Hz) were placed above the LCG on the left/right tibia, left/right femur, and sacrum of each participant. Five IMUs were also secured to the MIII spacesuit with tape and co-flex to the left/right upper leg, left/right lower leg, and hip brief. Custom sleeves at the hips and thighs were stitched into the LCG to add padding between the suited participant and MIII spacesuit. Foam padding (Viton) was inserted into these sleeves to alter the indexing between the participants and MIII. Volumetric scans were obtained at the U.S. Army Natick Army Center and ABF to obtain participant anthropometry and determine suit indexing (the level of padding added to the LCG) between the participant hips/thighs and MIII hip briefs. Participants performed a series of walking talks on an elevated walkway (10m long and 1m wide). For the unsuited condition, participants donned the TCG, LCG and human IMUs and performed 12 walking trails. For the suited condition, the MIII was pressurized to nominal suit pressure (4.3psi) in a tethered configuration (i.e. no closed-loop portable life support system (PLSS) was used). Participants donned the MIII with all 10 IMUs three times, each with different padding configurations at the hips and thighs: no (C0), single (C1), and double (C2) layer of padding. The approximate weight of the MIII without a human inside is 59kg; the actual total weight of the human and MIII varied based on the participant and configuration. For each suited condition, participants performed 24 walking trials. All 10 IMU sensors were wirelessly synchronized using manufacturer’s software at the beginning and end of each walking trial. In addition to walking talks, participants performed single and double leg balance tasks while unsuited and suited, but these data were not analyzed for this paper. The order in which unsuited and suited trials were performed was randomized and counterbalanced between all participants. For each step taken while the user was wearing the spacesuit, the following metrics were obtained: 1) step cadence, 2) human knee ROM (ROMH), 3) ΔROM, 4) ρ_HS^n (t) for the femur and tibia. The results from this study were initially presented at the 46th Annual International Conference on Environmental System and a journal paper publication is published in Aerospace Medicine and Human Performance journal. Following this study, this technology and methodology is being expanded to a wider range of tasks and analog environments. To collect data on a wider range of subjects with various suit sizes, this work will collaborate with research conducted during the EMU CO2 Washout study and the development of the subjective suit fit survey. This study was initially designed to quantify the carbon dioxide in the EMU spacesuit. This study is being led by Dr. Andrew Abercromby whom suggested this would be a good opportunity to test these suit fit metrics on a large sample size. Wearable sensor networks developed during the first year of this work have been added to the protocol of the CO2 washout study during which 20 subjects, with varying experience and fit within the EMU, will perform a series of cycle ergometer tests and work envelope exercises. While data is collected using these sensor networks, test engineers will guide test participants through an initial version of the subjective suit fit survey. Using data collected from these sensor networks, answers to subjective suit fit surveys, and outcome metrics collected for each task (i.e. completion time, metabolic load, etc.), initial analysis can begin to explore how suit fit can affect performance in operationally relevant scenarios and how well these metrics correlate with subjective feedback surveys. Preliminary results were presented at the 48th Annual International Conference on Environmental System.