Results for “Task complexity” — Page 27

NASA NTRSDate not supplied

Hybrid Model Based Approaches for Systems Health Management and Prognostics

To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Latest achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision making. In principle, data driven approaches can replace expensive experimental test-setups as well as reduce the number of simulations needed to explore, e.g., the parametric space of a multi-parameter model. Nonetheless, the limitations of pure data-driven methods came to light rather quickly, at least for some industries. In many industrial applications, data acquisition is costly, and the volume of data that can be collected does not satisfy the requirements for an effective model training and cross-validation. Therefore, some recent works in the area of machine learning is focusing on blending physics with data-driven algorithms, thus mitigating the drawbacks of the two approaches and emphasizing respective advantages. Partial physical knowledge of the problem can aid the learning process by “guiding” the algorithm towards efficient solutions that satisfy the physics driving the system behavior. The result is a hybrid modeling approach combining physical knowledge as well data driven methods to develop a unified hybrid approach. A hybrid framework for fusing information from physics-based performance models along with deep learning algorithms for prognostics of complex safety critical systems is presented. In this framework, physics-based performance models infer unobservable model parameters related to the system's components health solving a calibration problem in the deep learning approach.

Hybrid Modeling↗

NASA NTRSDate not supplied

Systems Health Management and Prognostics Approaches for Electric Aircrafts

As more and more electric vehicles emerge in our daily operation progressively, a very critical challenge lies in the prediction of remaining driving flying time/distance for the flying vehicles. This information is important, particularly in the case of auto vehicles, because such vehicles can become self-aware, autonomously compute its own capabilities, and identify how to best plan and successfully complete vehicular missions safely. In case of electric aircrafts, computing the remaining flying time is also safety-critical, since an aircraft that runs out of power (battery charge) while in the air will eventually lose control leading to catastrophe. To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Latest achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision-making. A systematic prediction framework is implemented to identify all possible sources of uncertainty, quantify each of them individually, and mathematically estimate their combined effect on the system-level quantity of interest, in this case, the remaining flying time/distance of the unmanned aircraft. Note - This presentation contains all previously published information.

Systems Health Managent↗

NASA NTRSDate not supplied

Evaluation of Self-Scheduling Exercises Completed by Analog Crewmembers in NASA's Human Exploration Research Analog (HERA)

NASA human spaceflight missions are inherently dynamic and require frequent scheduling changes in order to adapt to changing mission priorities and objectives. Tactical level changes to the mission plan are traditionally made by a team of expert planners and operations specialists on the ground. However, astronauts are expected to execute missions more autonomously during future long duration missions. Astronauts will need to take on some of the responsibility of managing their own schedule while still abiding by the numerous constraints required by human spaceflight operations. This paper summarizes salient elements of crew performance in NASA’s Human Exploration Research Analog Campaign 3. Analog crewmembers completed a series of self-scheduling exercises to evaluate Playbook’s usability towards enabling self-scheduling without support from ground control. Playbook is a self-scheduling software tool designed and developed by our team. We also investigated how to best communicate self-scheduling tasks and constraints to the crew in order to facilitate efficient self-scheduling during isolation in a realistic environment. Our analysis identified that 30 minutes was sufficient to complete complex self-scheduling tasks. Our evaluation also identified differences between individual and collaborative performance; analog crewmembers completed self-scheduling exercises more quickly as a team as opposed to individually and reported lower subjective difficulty ratings overall.

HERA↗

NASA NTRSDate not supplied

Hybrid Approaches to Systems Health Management and Prognostics

To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Latest achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision making. In principle, data driven approaches can replace expensive experimental test-setups as well as reduce the number of simulations needed to explore, e.g., the parametric space of a multi-parameter model. Nonetheless, the limitations of pure data-driven methods came to light rather quickly, at least for some industries. In many industrial applications, data acquisition is costly, and the volume of data that can be collected does not satisfy the requirements for an effective model training and cross-validation. Therefore, some recent works in the area of machine learning is focusing on blending physics with data-driven algorithms, thus mitigating the drawbacks of the two approaches and emphasizing respective advantages. Partial physical knowledge of the problem can aid the learning process by “guiding” the algorithm towards efficient solutions that satisfy the physics driving the system behavior. The result is a hybrid modeling approach combining physical knowledge as well data driven methods to develop a unified hybrid approach. A hybrid framework for fusing information from physics-based performance models along with deep learning algorithms for prognostics of complex safety critical systems is presented. In this framework, physics-based performance models infer unobservable model parameters related to the system's components health solving a calibration problem in the deep learning approach.

Systems Health Management↗

NASA NTRS2022

Hybrid Model Based Approaches for Systems Health Management and Prognostics

This is a previously approved and published presentation. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Present achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision-making. In principle, data-driven approaches can replace expensive experimental test-setups as well as reduce the number of simulations needed to explore, e.g., the parametric space of a multi-parameter model. Nonetheless, the limitations of pure data-driven methods came to light rather quickly, at least for some industries. In many industrial applications, data acquisition is costly, and the volume of data that can be collected does not satisfy the requirements for effective model training and cross-validation. Therefore, some recent works in the area of machine learning is focusing on blending physics with data-driven algorithms, thus mitigating the drawbacks of the two approaches and emphasizing respective advantages. Partial physical knowledge of the problem can aid the learning process by “guiding” the algorithm towards efficient solutions that satisfy the physics driving the system behavior. The result is a hybrid modeling approach combining physical knowledge as well data-driven methods to develop a unified hybrid approach. A hybrid framework for fusing information from physics-based performance models along with deep learning algorithms for prognostics of complex safety-critical systems is presented. In this framework, physics-based performance models infer unobservable model parameters related to the system's components health solving a calibration problem in the deep learning approach.

Prognostics↗

NASA NTRSDate not supplied

Health Monitoring and Prognostics for Electric Aircrafts

As more and more electric vehicles emerge in our daily operation progressively, a very critical challenge lies in the prediction of remaining driving flying time/distance for the flying vehicles. This information is important, particularly in the case of auto vehicles, because such vehicles can become self-aware, autonomously compute its own capabilities, and identify how to best plan and successfully complete vehicular missions safely. In case of electric aircrafts, computing the remaining flying time is also safety-critical, since an aircraft that runs out of power (battery charge) while in the air will eventually lose control leading to catastrophe. To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Latest achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision-making. A systematic prediction framework is implemented to identify all possible sources of uncertainty, quantify each of them individually, and mathematically estimate their combined effect on the system-level quantity of interest, in this case, the remaining flying time/distance of the unmanned aircraft. Note - This presentation contains all previously approved and published information.

Systems Health Managent↗

NASA NTRSDate not supplied

Physics Informed Neural Nets for Systems Health Management

To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Development in data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. The research work presents application of physics-informed neural nets application to a representative electric powertrain for unmanned aerial vehicles. The model is composed of physics-derived and empirical equations, integrated with connected networks that are strategically placed within the model to substitute equations that are subject to large uncertainty. Polynomial fit driven by heuristics or empirical observations can be substituted by more flexible networks that can minimize the error between model predictions and observations without being restricted to a predefined functional form. This modeling strategy allows training of networks deep inside the model and unknown parameters in a single learning stage.

Physics Informed↗

NASA NTRS1987

SOLON: An autonomous vehicle mission planner

The State-Operator Logic Machine (SOLON) Planner provides an architecture for effective real-time planning and replanning for an autonomous vehicle. The highlights of the system, which distinguish it from other AI-based planners that have been designed previously, are its hybrid application of state-driven control architecture and the use of both schematic representations and logic programming for the management of its knowledge base. SOLON is designed to provide multiple levels of planning for a single autonomous vehicle which is supplied with a skeletal, partially-specified mission plan at the outset of the vehicle's operations. This mission plan consists of a set of objectives, each of which will be decomposable by the planner into tasks. These tasks are themselves comparatively complex sets of actions which are executable by a conventional real-time control system which does not perform planning but which is capable of making adjustments or modifications to the provided tasks according to constraints and tolerances provided by the Planner. The current implementation of the SOLON is in the form of a real-time simulation of the Planner module of an Intelligent Vehicle Controller (IVC) on-board an autonomous underwater vehicle (AUV). The simulation is embedded within a larger simulator environment known as ICDS (Intelligent Controller Development System) operating on a Symbolics 3645/75 computer.

Dudziak, M. J.↗

NASA NTRS1993

Toward an embedded training tool for Deep Space Network operations

There are three issues to consider when building an embedded training system for a task domain involving the operation of complex equipment: (1) how skill is acquired in the task domain; (2) how the training system should be designed to assist in the acquisition of the skill, and more specifically, how an intelligent tutor could aid in learning; and (3) whether it is feasible to incorporate the resulting training system into the operational environment. This paper describes how these issues have been addressed in a prototype training system that was developed for operations in NASA's Deep Space Network (DSN). The first two issues were addressed by building an executable cognitive model of problem solving and skill acquisition of the task domain and then using the model to design an intelligent tutor. The cognitive model was developed in Soar for the DSN's Link Monitor and Control (LMC) system; it led to several insights about learning in the task domain that were used to design an intelligent tutor called REACT that implements a method called 'impasse-driven tutoring'. REACT is one component of the LMC training system, which also includes a communications link simulator and a graphical user interface. A pilot study of the LMC training system indicates that REACT shows promise as an effective way for helping operators to quickly acquire expert skills.

Hill, Randall W., Jr.↗

NASA NTRS1993

Toward an Embedded Training Tool for Deep Space Network Operations

There are three issues to consider when building an embedded training system for a task domain involving the operation of complex equipment: (1) how skill is acquired in the task domain; (2) how the training system should be designed to assist in the acquisition of the skill, and more specifically, how an intelligent tutor could aid in learning; and (3) whether it is feasible to incorporate the resulting training system into the operational environment. This paper describes how these issues have been addressed in a prototype training system that was developed for operations in NASA's Deep Space Network (DSN). The first two issues were addressed by building an executable cognitive model of problem solving and skill acquisition of the task domain and then using the model to design an intelligent tutor.

Johnson, W. Lewis↗

NASA NTRS1992

Design concepts for the development of cooperative problem-solving systems

There are many problem-solving tasks that are too complex to fully automate given the current state of technology. Nevertheless, significant improvements in overall system performance could result from the introduction of well-designed computer aids. We have been studying the development of cognitive tools for one such problem-solving task, enroute flight path planning for commercial airlines. Our goal was two-fold. First, we were developing specific systems designs to help with this important practical problem. Second, we are using this context to explore general design concepts to guide in the development of cooperative problem-solving systems. These designs concepts are described.

Smith, Philip J.↗

NASA NTRS1995

Enroute flight planning: Evaluating design concepts for the development of cooperative problem-solving systems

There are many problem-solving tasks that are too complex to fully automate given the current state of technology. Nevertheless, significant improvements in overall system performance could result from the introduction of well-designed computer aids. We have been studying the development of cognitive tools for one such problem-solving task, enroute flight path planning for commercial airlines. Our goal has been two-fold. First, we have been developing specific system designs to help with this important practical problem. Second, we have been using this context to explore general design concepts to guide in the development of cooperative problem-solving systems. These design concepts are described below, along with illustrations of their application.

Smith, Philip J.↗

DOE OSTI2026

Generative large language models for predictive maintenance planning

Maintenance planning and the generation of necessary components for tasks can prove time-consuming and complex. Automating the creation of recurring or similar tasks by leveraging previous planning packages and data, while uncovering insights to automate planning package generation, presents an opportunity to conserve valuable time and resources. This work aims to harness the textual and probabilistic capabilities of large language models (LLMs) to automate the generation of planning packages. Utilizing diverse data sources ranging from raw data to handwritten text, both singular and collaborative LLMs are trained and tested. Results demonstrate their capability to generate essential planning package components, effectively replicating the statistical patterns in the data. This demonstrates the use of these tools inside a digital asset for automated planning. This work outlines a methodology for constructing datasets, a training suite, and evaluation methods for LLM-based textual and conversational planning tools utilized in an asset digital twin. Results indicate that the fine-tuned models generate estimated planning information within the statistical ranges observed in real maintenance data. The models achieve high accuracy (>90%) in document question-answering and instruction generation tasks. Furthermore, the conversational retrieval-augmented generation (RAG) assistant system achieves 100% document retrieval accuracy, while conversational information capture exceeds 98% across the majority of work-package assistant modules.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1991

Expert Script Generator

Program provides additional level of interface to facilitate use of telerobotic system. ESG (Expert Script Generator) is software package automatically generating high-level task objective commands from complex menu-driven language of the NASA Intelligent Systems Research Laboratory (ISRL). Makes telerobotics laboratory accessible to researchers not familiar with comprehensive language developed by ISRL for interacting with various systems of ISRL test bed. Incorporates expert-system technology to capture typical rules of operation that skilled operator uses. Result: operator interfact optimizing ability of system to perform task remotely in hazardous environment, in timely manner, and without undue stress to operator, while minimizing change for operator erros that damage equipment. Written in CLIPS.

Sliwa, Nancy E.↗

DOE OSTI2025

El Agente: An autonomous agent for quantum chemistry

Computational chemistry tools are widely used to study the behavior of chemical phenomena. Yet, the complexity of these tools can make them inaccessible to non-specialists and challenging even for experts. In this work, we introduce El Agente Q, an LLM-based multi-agent system that dynamically generates and executes quantum chemistry workflows from natural language user prompts. The system is built on a novel cognitive architecture featuring a hierarchical memory framework that enables flexible task decomposition, adaptive tool selection, post-analysis, and autonomous file handling and submission. El Agente Q is benchmarked on six university-level course exercises and two case studies, demonstrating robust problem-solving performance (averaging >87% task success) and adaptive error handling through in situ debugging. It also supports longer-term, multi-step task execution for more complex workflows, while maintaining transparency through detailed action trace logs. Together, these capabilities lay the foundation for increasingly autonomous and accessible quantum chemistry.

agentic systems↗

NASA NTRSDate not supplied

Scenario-Based Task Design for Airline Pilot Anticipatory Behaviors: Asynchronous Assessment of Complex Cognitive Skills

Airline pilots must anticipate potential threats to passenger safety and efficient flight. Such anticipation needs to occur at both the system and individual levels, yet no formal training for these anticipatory behaviors currently exists. The emerging field of resilience engineering provides a framework to explore this problem. Given the complex nature of this problem, the assessment portion of the project constituted a challenge. Simulator time is expensive as simulators are costly to run, and they are extremely limited in number. Also, by the nature of their work, pilots as a group are hard to connect with in person; thus, asynchronous methods were adopted. Further, to ensure that authentic pilot behaviors were observed, task-based scenarios were used rather than focus group or interview protocols. This session explores the design of asynchronous learning and a measure for assessing pilot anticipatory behaviors outside of a simulator to help design learning opportunities for pilots. The scenario-based and active assessment items will be discussed in detail, and demonstrations of these items will be shared. These items constitute the bulk of the challenge for this work. At the time of this writing, data is being collected to assess the effectiveness of the interactive training module and validate the measures. This data will be analyzed and presented as part of the session. This will include any linkages between the scenario-based and traditional assessment items.

task design↗

DOE OSTI2024

Decode the Workload: Training Deep Learning Models for Efficient Compute Cluster Representation

Monitoring the status of a high throughput computing cluster running computationally intensive production jobs is a crucial yet challenging system administration task due to the complexity of such systems. To this end, we train autoencoders using the Linux kernel CPU metrics of the cluster. Additionally, we explore assisting these models with graph neural networks to share information across threads within a compute node. The models are compared in terms of their ability to: 1) Produce a compressed latent representation that captures the salient features of the input, 2) Detect anomalous activity, and 3) Make distinction between different kinds of jobs run at Jefferson Lab. The goal is to have a robust encoder whose compressed embeddings are used for several downstream tasks. We extend this study further by deploying these models in a human-in-the-loop production-based setting for the anomaly detection task and discuss the associated implementation aspects such as continual learning and the criterion to generate alarms. This study represents a first step in the endeavor towards building self-supervised large-scale foundation models for computing centers.

Mohammed, Ahmed↗

NASA NTRS1975

Secondary task for full flight simulation incorporating tasks that commonly cause pilot error: Time estimation

The task of time estimation, an activity occasionally performed by pilots during actual flight, was investigated with the objective of providing human factors investigators with an unobtrusive and minimally loading additional task that is sensitive to differences in flying conditions and flight instrumentation associated with the main task of piloting an aircraft simulator. Previous research indicated that the duration and consistency of time estimates is associated with the cognitive, perceptual, and motor loads imposed by concurrent simple tasks. The relationships between the length and variability of time estimates and concurrent task variables under a more complex situation involving simulated flight were clarified. The wrap-around effect with respect to baseline duration, a consequence of mode switching at intermediate levels of concurrent task distraction, should contribute substantially to estimate variability and have a complex effect on the shape of the resulting distribution of estimates.

Rosch, E.↗