Search NASA⌕ Search

SEARCH · Search NASA

Results for “task scenario”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 163 records · Page 9

Validation of Fitness for Duty Standards Using Pre- and Post-Flight Capsule Egress and Suited Functional Performance Tasks in Simulated Reduced Gravity

The transition between gravity environments will involve one of the most complex, high-risk phases of any mission. For example, the reduced functional capacity caused by physiological deconditioning adaptations in microgravity coupled with the stressors of re-entry into partial gravity environments will increase risks to crew, even with rigorous adherence to inflight countermeasures. Quantification of the astronauts’ post-landing functional performance is necessary to design ConOps for exploration missions. Specifically, these two high-risk scenarios may be required to be performed soon after gravity transitions: • Nominal and/or emergency unassisted capsule egress task after return to Earth • Planetary extravehicular activity (EVA) soon after landing on Mars or the Moon This study is broken down into two phases. Phase 1: A feasibility study that will assess the overall feasibility and demonstrate the capability to do these tasks shortly after landing. Phase 2: the full Egress Fitness study, which is part of the CIPHER complement. This study uses a task-based approach to characterize functional performance of these high-risk scenarios in long-duration ISS crewmembers before flight and shortly after return to Earth. Prior to any testing, each astronaut subject completes a suit fit check to ensure adequate sizing and a mobility assessment to confirm completion of the EVA tasks. Pilot Egress Fitness pre-flight and post-flight testing includes an Earth-based emergency egress out of a functional capsule mockup and a short Mars gravity EVA simulation including suit donning, hatch egress, ladder descent, task board cable operations, baggage transfer over sand/rocky regolith, alignment with a rear entry port, and suit egress. Pre-flight testing will occur at any time point prior to flight. Post-flight assessment is much more critical with the capsule egress test that will occur 1–4 hours after landing and the planetary EVA approximately 18–36 hours after landing. The CIPHER study will incorporate additional pre-flight sessions, longer EVA tasks that include traverse and geology sampling, and post-flight sessions on R+1, 4, and 8 to characterize the timeframe of recovery. Data collected for both tasks include task completion time, photo, and video. The EVA portion also includes collection of metabolic and heart rate. Pilot Egress Fitness study has completed both baseline and post-flight testing on four astronaut subjects, with all four able to complete post-flight EVA testing and three able to complete postflight capsule egress testing. Preliminary results observe individual physiological variations, which were to be expected but also suggest the need to carefully track the timeline from undock to landing to testing. Furthermore, some task performance instructions and equipment may need to be adjusted to ensure results are primarily physiological. These and other lessons learned will be addressed with minor protocol changes in the full CIPHER Egress Fitness study.

J R Norcross↗

Supervisory Control with Dual Tasking in Post G-Transition Vehicle Landings

BACKGROUND Landing during exploration spaceflight may consist of both planned automated supervisory control and unplanned crew override. Supervisory control, particularly when performed under cognitive load with additional monitoring tasks, is essential for ensuring overall mission success during landing contingencies. Evaluating performance in a relevant Human Landing System (HLS) supervisory landing task after long-duration microgravity exposure can help identify potential risks from human error and sensorimotor alterations. Adaptive changes in the sensorimotor system can manifest during g-transitions as spatial disorientation. Although training and landing aids facilitate successful landings despite disorientation, these adaptive changes may heighten cognitive demand, which must be considered in the landing strategy. It is important to characterize these effects as soon as possible following the G-transition while the sensorimotor system remains in a state of adaptive flux to inform appropriate countermeasures. METHODS A Multi-attribute Lunar Table Battery (MALTB) task was developed for an iOS tablet device to provide flexible crew testing and training capabilities in-flight and on the ground. Elements of the tablet task were derived from the Multi-Attribute Task Battery (MATB, Cegarra et al. 2020). The task requires crew members to study a map of a planned landing site and memorizing the terrain and surface landmarks to inform potential divert maneuvers during landing. The user will oversee a series of approaches through touchdown simulations on the tablet with an external camera view of the Lunar surface. The primary responsibility of the crew member will be to execute a divert if the guidance recommended site is erroneous (e.g., the guidance projected landing target is not within 10m of the planned landing site center), or the projected landing site is no longer suitable due to surface obstacles. Considering vehicle maneuverability and fuel reserves, the divert capabilities will diminish as the task progresses. In cases where a divert is initiated, a new landing target will need to be designated by the user and will be evaluated for the proximity to the original pre-planned site. A secondary operational monitoring task will challenge the user's cognitive reserve by requiring the user to maintain several gauges within acceptable limits and respond to a visual indicator while completing the landing approach. Outcome measures include distance from the planned landing site to the user-initiate divert landing location, ground slope at the new landing site, time to divert, the ability to accomplish the secondary monitoring tasks, and perceived workload. The tablet task is being evaluated in a ground-based study to determine the learning effect of first-time users. The tablet task will be utilized in a flight study to test performance multiple times postflight. Future potential testing in-flight have been identified for capsules with iOS tablet devices. Multi-attribute Lunar Table Battery Task MALTB comprises video footage of thirty distinct landing conditions. The landing scenarios feature various landing sites (n = 3), hazardous object sizes (n = 6), sun azimuth degrees (n = 4), camera modes (i.e., fixed or gimbaled), and navigation bias (i.e., true or false). Each seventy second trial uses a sixty-degree constant glideslope trajectory. The application architecture and layout include user identification setup and data storing, a guided walkthrough of the task and interface components, practice mode for task familiarization, and a modified Bedford workload scale administered following task completion. The time-based dependent measures are saved locally to the iOS Files application and post-processing scripts have been developed to evaluate the remaining measures of performance. RELEVANCE This project will deliver an operational demonstration of crew monitoring capability following spaceflight and identify potential deficits that may require remediation. Comparison of individual vestibular and cognitive changes with crew performance will help better characterize the landing risks associated with sensorimotor alterations. ACKNOWLEDGEMENTS: This project is funded by NASA’s Human Research Program Human Health Countermeasures Element. REFERENCES Cegarra J, Valery B, Avril E, Calmettes C, Navarro J (2020) OpenMATB: A Multi-Attribute Task Battery promoting task customization, software extensibility and experiment replicability. Behav Res Methods 52:1980-1990 doi: 10.3758/s13428-020-01364-w

Matthew McDonnell↗

Testing, Calibration, and UxS Integration of the Kromek GR1 Plus CZT Gamma Spectrometer

Collecting radiation measurements can be a hazardous task, especially in the presence of highly active sources. Various scenarios necessitate source search, classification, and quantification, often with limited or no a priori information. Activities such as disaster mitigation, emergency response, environmental monitoring, and site remediation may involve dangerous radioactive sources. In these situations, there is a pressing need for remote monitoring capabilities that protect human operators from potential harm and enhance adherence to the principle of “As Low as Reasonably Achievable” (ALARA) for radiation doses. The first step toward achieving remote radiation measurement capabilities is the remote operation of a radiation sensor. Once this milestone is reached, the next challenge is to integrate this remote sensing capability into suitable actuation agents, collectively referred to as uncrewed systems (UxS). These systems include familiar platforms such as robotic quadrupeds, aerial multirotor vehicles, and ground vehicles, any of which may be teleoperated, act autonomously, or utilize a combination of both. A critical factor in achieving remote radiation sensing is the availability of data from the appropriate sensor. Many commercially available radiation sensors have closed-source documentation for their communication protocols. Typical end-user products are often self-contained, handheld devices designed for manual measurement scenarios. While there are commercial off-the-shelf (COTS) integrations of radiation sensors with UxS available for purchase, these solutions are typically tailored for specific use cases and may not meet the requirements of different applications. This paper discusses efforts to remotely acquire radiation measurements from a small form-factor CZT gamma spectrometer. Sandia National Laboratories has successfully demonstrated the initial capability to integrate low size, weight, and power (SWaP) gamma spectroscopy into various UxS, alongside co-located GPS data logging and sensor calibration and qualification. With remote gamma spectroscopy achieved, the stage is set for UxS integration of this capability.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Structural assembly in space

A cost algorithm for predicting assembly costs for large space structures is given. Assembly scenarios are summarized which describe the erection, deployment, and fabrication tasks for five large space structures. The major activities that impact total costs for structure assembly from launch through deployment and assembly to scientific instrument installation and checkout are described. Individual cost elements such as assembly fixtures, handrails, or remote minipulators are also presented.

Stokes, J. W.↗

Crew procedures and workload of retrofit concepts for microwave landing system

Crew procedures and workload for Microwave Landing Systems (MLS) that could be retrofitted into existing transport aircraft were evaluated. Two MLS receiver concepts were developed. One is capable of capturing a runway centerline and the other is capable of capturing a segmented approach path. Crew procedures were identified and crew task analyses were performed using each concept. Crew workload comparisons were made between the MLS concepts and an ILS baseline using a task-timeline workload model. Workload indexes were obtained for each scenario. The results showed that workload was comparable to the ILS baseline for the MLS centerline capture concept, but significantly higher for the segmented path capture concept.

Summers, Leland G.↗

An Autonomous Control System for an Intra-Vehicular Spacecraft Mobile Monitor Prototype

This paper presents an overview of an ongoing research and development effort at the NASA Ames Research Center to create an autonomous control system for an internal spacecraft autonomous mobile monitor. It primary functions are to provide crew support and perform intra- vehicular sensing activities by autonomously navigating onboard the International Space Station. We describe the mission roles and high-level functional requirements for an autonomous mobile monitor. The mobile monitor prototypes, of which two are operational and one is actively being designed, physical test facilities used to perform ground testing, including a 3D micro-gravity test facility, and simulators are briefly described. We provide an overview of the autonomy framework and describe each of its components, including those used for automated planning, goal-oriented task execution, diagnosis, and fault recovery. A sample mission test scenario is also described.

Dorais, Gregory A.↗

Examining the Relationship between Self-Reported Workload and Resilient Performance in Airline Flight Crews

Workload has long been associated with human performance in aviation. High workload is typically viewed as a system design problem. The aim of this work was to find observable behaviors that guard cognitive resource margins; thus, improve handling of perturbations that arise. Twelve commercial airline flight crews participated in an experiment at NASA Langley Research Center. We explored workload data (NASA Task Load Index) self-reported after completing flight simulation scenarios to see how different individuals, using the same technologies, cope with challenging situations. A trained observer from the same airline as the study participants assessed performance. We differentiated a lower workload crew from a higher workload crew and analyzed their workload and performance. Results indicate that the workload may be a component of pilots’ resilient performance. Initial strategies that reduced workload were classified and these types of behaviors, if implemented correctly, might assist pilots in maintaining functional workload resource margins.

Michael Stewart↗

Examining the Relationship Between Workload and Resilient Performance in Airline Flight Crews

Workload has long been associated with human performance in aviation. High workload is typically viewed as a system design problem. The aim of this work was to find observable behaviors that guard cognitive resource margins; thus, improve handling of perturbations that arise. Twelve commercial airline flight crews participated in an experiment at NASA Langley Research Center. We explored workload data (NASA Task Load Index) self-reported after completing flight simulation scenarios to see how different individuals, using the same technologies, cope with challenging situations. A trained observer from the same airline as the study participants assessed performance. We differentiated a lower workload crew from a higher workload crew and analyzed their workload and performance. Results indicate that the workload may be a component of pilots’ resilient performance. Initial strategies that reduced workload were classified and these types of behaviors, if implemented correctly, might assist pilots in maintaining functional workload resource margins.

Michael Stewart↗

Effects of Modality on Interrupted Flight Deck Performance: Implications for Data Link

Externally-imposed tasks frequently interrupt ongoing task performance in the commercial flight deck. While normally managed without consequence, basic research as well as aviation accident and incident investigations show that interruptions can negatively affect performance and safety. This research investigates the influence of interruption and interrupted task modality on pilot performance in a simulated commercial flight deck. Fourteen current commercial airline pilots performed approach scenarios in a fixed-base flight simulator. Air traffic control instructions, conveyed either aurally or visually (via a data link system) interrupted a visual task (obtaining information from the Flight Management System) and an auditory task (listening to the automated terminal information service recording). Some results confirm the hypothesized performance advantage of cross-modality conditions, more compelling nature of auditory interruptions, and interruption-resistance of auditory ongoing tasks. However, taken together, results suggest the four interaction conditions had different effects on pilot performance. These results have implications for the design of data link systems, and for facilitating interruption management through interface design, aiding, and training programs.

Latorella, Kara A.↗

Multiphysics Modeling of Microreactors with NEAMS codes, and Validation Based on KRUSTY Reactivity Insertion

The NEAMS Multiphysics Applications team continues to assess code usability and functionality for microreactor design and safety analyses, while demonstrating that NEAMS tools capture both steady-state and transient behavior across distinct microreactor concepts. In FY2025, the team advanced full-core, high-fidelity, multiphysics models that solve more complex problems and strengthen verification/validation for several microreactor systems: heat-pipe microreactor (HPMR), gas-cooled microreactor (GCMR), and the KRUSTY experiment. These models employ the MOOSE MultiApp/Transfers architecture with Griffin for neutronics, BISON for heat conduction/thermomechanics, Sockeye for heat pipes, SAM/THM for coolant channels and loops, and SWIFT for hydride behavior, with meshes generated via the MOOSE Reactor Module. The graphite models available in the Grizzly code were also investigated for future analyses. For the HPMR, a Na-HPMR variant was constructed to align with recently validated heat-pipe experiments and Sockeye’s LCVF capability, enabling mechanistic heat-pipe transients and startup modeling. The Na-HPMR will serve as the primary model for HPMR investigations in upcoming tasks. The load-following and single heat-pipe failure scenarios (Griffin/BISON/Sockeye), which were previously modeled for the K-HPMR, were replicated for the Na-HPMR, showing strong negative temperature feedback and highly localized thermal effects, respectively, while the startup case captured vapor-front progression and heat-removal activation. Solid mechanics was added to the previously built K-HPMR full-core model in BISON, showing minimal impact on steady-state reactivity yet enabling stress-field predictions that prepare the path for full-core TRISO performance analyses. For the GCMR, automated steady-state and four transient scenarios were executed using Griffin/BISON/SAM/SWIFT. Results confirm robust inherent safety: power collapses promptly in loss-of-cooling events, the inlet-temperature drop settles to a new equilibrium, and a single-channel blockage yields only a ~30 K local fuel-temperature rise with <0.4% power decrease. SWIFT-predicted hydrogen redistribution affects reactivity during both steady-state and transient conditions, underscoring its importance. A Brayton-cycle balance of plant (BOP) model in SAM/THM demonstrated stable startup behavior, and xenon-driven reactivity during load following was analyzed. To improve TRISO-compact temperature fidelity, a fast multiscale Heat Source Decomposition (HSD) treatment was implemented. Against heterogeneous benchmarks, HSD reduces underprediction of kernel temperatures and lowers predicted peak powers in reactivity-insertion transients compared to previous homogenized models. KRUSTY warm-critical validation progressed from FY2024 baselines: the 15Ȼ insertion shows excellent agreement in peak power (~2% high) and temperature trends, and the 30Ȼ case was automated via a feedback controller that maintained power near 3 kW for ~150 s with close agreement to data. The successful modeling of the warm critical tests has laid a strong foundation for simulating more complex nuclear system tests in the years ahead. Throughout FY2025, developer feedback was provided (e.g., MOOSE batch mesh generation, distributed pre-split meshes, Griffin sweeper on displaced meshes), several new models were contributed to the Virtual Test Bed, and an OECD-NEA WPRS multiphysics benchmark based on the HPMR was initiated to enable broader cross-comparison and best-practice development with the nuclear community at large.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

A space transportation system operations model

Presented is a description of a computer program which permits assessment of the operational support requirements of space transportation systems functioning in both a ground- and space-based environment. The scenario depicted provides for the delivery of payloads from Earth to a space station and beyond using upper stages based at the station. Model results are scenario dependent and rely on the input definitions of delivery requirements, task times, and available resources. Output is in terms of flight rate capabilities, resource requirements, and facility utilization. A general program description, program listing, input requirements, and sample output are included.

Morris, W. Douglas↗

Army-NASA aircrew/aircraft integration program (A3I) software detailed design document, phase 3

The capabilities and design approach of the MIDAS (Man-machine Integration Design and Analysis System) computer-aided engineering (CAE) workstation under development by the Army-NASA Aircrew/Aircraft Integration Program is detailed. This workstation uses graphic, symbolic, and numeric prototyping tools and human performance models as part of an integrated design/analysis environment for crewstation human engineering. Developed incrementally, the requirements and design for Phase 3 (Dec. 1987 to Jun. 1989) are described. Software tools/models developed or significantly modified during this phase included: an interactive 3-D graphic cockpit design editor; multiple-perspective graphic views to observe simulation scenarios; symbolic methods to model the mission decomposition, equipment functions, pilot tasking and loading, as well as control the simulation; a 3-D dynamic anthropometric model; an intermachine communications package; and a training assessment component. These components were successfully used during Phase 3 to demonstrate the complex interactions and human engineering findings involved with a proposed cockpit communications design change in a simulated AH-64A Apache helicopter/mission that maps to empirical data from a similar study and AH-1 Cobra flight test.

Banda, Carolyn↗

Experimental Studies of the Effect of Intent Information on Cockpit Traffic Displays

Intent information provides knowledge of another aircraft's current and future trajectory states. Prototype traffic displays were designed for four different levels of intent: Position, Rate, Commanded State, and FMS (Flight Management System)-Path. The current TCAS (traffic collision avoidance systems) Display, which shows altitude rate in addition to current position and altitude, was used as a baseline and represents the lowest level of intent. The Rate, Commanded State, and FMS-Path Displays show increasing levels of intent information using TCAS-like symbology in addition to incorporating a conflict probe and profile view display. An initial experiment was run on the MIT (Massachusetts Institute of Technology) Part Task Flight Simulator in which eight airline pilots flew five traffic scenarios with each of the four displays. Results show that pilots had fewer separation violations and maneuvered earlier with the three intent displays. Separation violations were reduced when pilots maneuvered earlier. A second experiment was run to compare performance between displaying intent information directly and incorporating it into a conflict probe. A different set of eight airline pilots flew four traffic scenarios with the TCAS and Commanded State Displays with and without the conflict probe. Conflict probes with two minute and long range look-ahead times were tested. Displaying conflict bands or showing intent information directly both led to fewer separation violations and earlier avoidance maneuvers than the base TCAS Display. Performance was similar between the two minute and long range look-ahead conflict probes. Pilots preferred all intent displays over the TCAS Display.

Barhydt, Richard↗

Experimental Studies of Intent Information on Cockpit Traffic Displays

Intent information provides knowledge of another aircraft's current and future trajectory states. Prototype traffic displays were designed for four different levels of intent: No Intent, Rate, Commanded State, and Flight Management System (FMS)-Path. The TCAS Display was used as a baseline and represents the No Intent Level. The Rate, Commanded State, and FMS-Path Displays show increasing levels of intent information using TCAS-like symbology in addition to incorporating a conflict probe and profile view display. An experiment was run on the MIT Part Task Flight Simulator in which eight airline pilots flew five traffic scenarios with each of the four displays. Results show that pilots had fewer separation violations and maneuvered earlier with the three intent displays. Separation violations were reduced when pilots maneuvered earlier. A second experiment was run to compare performance between displaying intent information directly and incorporating it into a conflict probe. A different set of eight airline pilots flew four traffic scenarios with the TCAS and Commanded State Displays with and without the conflict probe. Conflict probes with two minute and long range look-ahead times were tested. Displaying conflict bands or showing intent information directly both led to fewer separation violations and earlier avoidance maneuvers than the base TCAS Display. Performance was similar between the two minute and long range look-ahead conflict probes. Pilots preferred all intent displays over the TCAS Display.

Barhydt, Richard↗

Development of Improved Models, Stochasticity, and Frameworks for the MIT Extensible Air Network Simulation

MEANS, the MIT Extensible Air Network Simulation, was created in February of 2001, and has been developed with support from NASA Ames since August of 2001. MEANS is a simulation tool which is designed to maximize fidelity without requiring data of such a low level as to preclude easy examination of alternative scenarios. To this end, MEANS is structured in a modular fashion to allow more detailed components to be brought in when desired, and left out when they would only be an impediment. Traditionally, one of the difficulties with high-fidelity models is that they require a level of detail in their data that is difficult to obtain. For analysis of past scenarios, the required data may not have been collected, or may be considered proprietary and thus difficult for independent researchers to obtain. For hypothetical scenarios, generation of the data is sufficiently difficult to be a task in and of itself. Often, simulations designed by a researcher will model exactly one element of the problem well and in detail, while assuming away other parts of the problem which are not of interest or for which data is not available. While these models are useful for working with the task at hand, they are very often not applicable to future problems. The MEAN Simulation attempts to address these problems by using a modular design which provides components of varying fidelity for each aspect of the simulation. This allows for the most accurate model for which data is available to be used. It also provides for easy analysis of sensitivity to data accuracy. This can be particularly useful in the case where accurate data is available for some subset of the situations that are to be considered. Furthermore, the ability to use the same model while examining effects on different parts of a system reduces the time spent learning the simulation, and provides for easier comparisons between changes to different parts of the system.

Clarke, John-Paul↗

Sensor Web Interoperability Testbed Results Incorporating Earth Observation Satellites

This paper describes an Earth Observation Sensor Web scenario based on the Open Geospatial Consortium s Sensor Web Enablement and Web Services interoperability standards. The scenario demonstrates the application of standards in describing, discovering, accessing and tasking satellites and groundbased sensor installations in a sequence of analysis activities that deliver information required by decision makers in response to national, regional or local emergencies.

Frye, Stuart↗

Evaluation of Synthetic Vision Display Concepts for Improved Awareness in Unusual Attitude Recovery Scenarios

A recent study conducted by the Commercial Aviation Safety Team (CAST) determined 40 percent of all fixed-wing fatal accidents, between 2001 and 2011, were caused by Loss-of-Control (LOC) in flight (National Transportation Safety Board, 2015). Based on their findings, CAST recommended manufacturers develop and implement virtual day-visual meteorological conditions (VMC) display systems, such as synthetic vision or equivalent systems (CAST, 2016). In a 2015 simulation study conducted at NASA Langley Research Center (LaRC), researchers gathered to test and evaluate virtual day-VMC displays under realistic flight operation scenarios capable of inducing reduced attention states in pilots. Each display concept was evaluated to determine its efficacy to improve attitude awareness. During the experiment, Evaluation Pilots (EPs) were shown the following three display concepts on the Primary Flight Display (PFD): Baseline, Synthetic Vision (SV) with color gradient, and SV with texture. The baseline configuration was a standard, conventional 'blue over brown' display. Experiment scenarios were simulated over water to evaluate Unusual Attitude (UA) recovery over 'featureless terrain' environments. Thus, the SV with color gradient configuration presented a 'blue over blue' display with a linear blue color progression, to differentiate attitude changes between sky and ocean. The SV with texture configuration presented a 'blue over blue' display with a black checkerboard texture atop a synthetic ocean. These displays were paired with a Background Attitude Indicator (BAI) concept. The BAI was presented across all four Head-Down Displays (HDDs), displaying a wide field-of-view blue-over-blue attitude indicator. The BAI aligned with the PFD and showed through the background of the navigation displays with opaque transparency. Each EP participated in a two-part experiment series with a total seventy-five trial runs: Part I included a set of twenty-five Unusual Attitude Recovery (UAR) scenarios; Part II included a set of fifty Attitude Memory Recall Tasks (AMRT). At the conclusion of each trial, EPs were asked to complete a set post-run questionnaires. Quantitative results showed that there were no significant statistical effects on UA recovery times when utilizing SV with or without the presence of a BAI. Qualitative results show the SV displays (color, texture) with BAI On are most preferred for both UA recognition and recovery when compared with the baseline display. When only comparing SV display concepts, EPs performed better when using the SV with texture, BAI On, than any other display configuration. This is an interesting find considering most EPs noted their preference towards the SV with color gradient when the BAI was on.

Nicholas, Stephanie↗

Improving Sim-to-Real Transfer in Vision-Based Robot Navigation Via Instance-Level GAN-Based Data Augmentation

Achieving robust vision-based robotic tasks requires large amounts of data, which are often difficult to obtain in real-world scenarios. Simulators and synthetic data offer a cost-effective alternative, but the visual gap between simulation and reality hinders the performance of models when deployed in real-world environments. In this paper, we present a data augmentation pipeline that integrates a foundation model (Segment Anything Model) with an unsupervised image-to-image translation model (CycleGAN) for instance-level domain transfer from simulation to reality. This pipeline enables the generation of realistic labeled data from synthetic images for training supervised machine learning models in vision-based navigation tasks. We evaluate our approach on real-world data for ego-vehicle pose estimation, a critical autonomous navigation task involving the prediction of cross-track position and heading angle relative to road center line markings. The results of our tests show that our GAN-based data augmentation pipeline significantly outperforms models trained solely on simulation data or on data processed with standard image augmentation methods for sim-to-real transfer, enhancing model robustness and generalizability in real-world scenarios. Our method provides a scalable and flexible data augmentation tool for leveraging large synthetic datasets to enhance vision-based robotic navigation tasks.

artificial intelligence↗