Remote manipulation with transmission delay.
Performance of simple and complex tasks affected by inserting transmission delay between master and slave elements of remote manipulator
OPERATOR PERFORMANCE↗
SEARCH · Search NASA
Results for “complex task performance”
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.
Physiological Based Simulator Fidelity Design Guidance
The evolution of the role of flight simulation has reinforced assumptions in aviation that the degree of realism in a simulation system directly correlates to the training benefit, i.e., more fidelity is always better. The construct of fidelity has several dimensions, including physical fidelity, functional fidelity, and cognitive fidelity. Interaction of different fidelity dimensions has an impact on trainee immersion, presence, and transfer of training. This paper discusses research results of a recent study that investigated if physiological-based methods could be used to determine the required level of simulator fidelity. Pilots performed a relatively complex flight task consisting of mission task elements of various levels of difficulty in a fixed base flight simulator and a real fighter jet trainer aircraft. Flight runs were performed using one forward visual channel of 40 deg. field of view for the lowest level of fidelity, 120 deg. field of view for the middle level of fidelity, and unrestricted field of view and full dynamic acceleration in the real airplane. Neuro-cognitive and physiological measures were collected under these conditions using the Cognitive Avionics Tool Set (CATS) and nonlinear closed form models for workload prediction were generated based on these data for the various mission task elements. One finding of the work described herein is that simple heart rate is a relatively good predictor of cognitive workload, even for short tasks with dynamic changes in cognitive loading. Additionally, we found that models that used a wide range of physiological and neuro-cognitive measures can further boost the accuracy of the workload prediction.
Schnell, Thomas↗
IRIS: Exploring Performance Scaling of the Intelligent Runtime System and its Dynamic Scheduling Policies
High-Performance Computing is becoming increasingly heterogeneous, relying on a diverse mix of hardware to achieve good performance. Paradoxically, current drivers and frameworks for these devices typically require separate languages and implementations for each vendor. Furthermore, there are few tools and little support to schedule codes between these devices in a truly heterogeneous manner-partly because of this fragmentation between vendors and the languages each supports. To overcome both limitations, the Intelligent Runtime System (IRIS) was developed. It allows a common task abstraction to automatically be shared among contemporary vendors and is run from a single host-side API. At runtime, IRIS queries the host system and registers which frameworks and drivers are available, these determine which kernels can be used by the scheduler-CPUs via OpenMP, Nvidia GPUs (CUDA), AMD GPUs (HIP), and Intel and Xilinx FPGAs with OpenCL. IRIS enables tasks to be scheduled to any heterogeneous device and resolves to the appropriate kernel binary at runtimeit only uses the devices supported by the system on which it is run. IRIS supports single-task and graph-based expressions of dependencies of tasks. Additionally, IRIS features a range of dynamic scheduling policies, allowing complex chains of tasks and interactions to be executed, relieving the programmer/user from considering the system to assign tasks to devices optimally. This paper presents the peak performance attainable by IRIS over a range of systems-each with different numbers and types of accelerator devices, it highlights the flexibility of IRIS since these devices are truly heterogeneous, relying on different backends (drivers, frameworks, and languages) which historically required unique implementations to utilize them. We then use this peak performance as a baseline to compare increasingly complex chains of tasks (with increasingly complex task dependencies) and evaluate how IRIS copes. Finally, we consider the performance of different IRIS scheduling policies on this range of task graphs.
Johnston, Beau↗
Synaptic Functionality and Neuromorphic Information Processing in Membrane Ion Channel Junctions
The human brain performs complex memory and computational tasks with high energy efficiency by regulating ion transport through membrane channels. These signaling mechanisms have been inspiring the development of nanofluidic memristors that emulate synaptic behavior. Here, in this study, we describe a membrane ion channel synapse (MICS), constructed from aqueous droplets linked by gramicidin A channels, that achieves neuromorphic functionality. MICS exhibits memristive ion transport with hysteretic current–voltage behavior arising from voltage-dependent channel formation and ion transport dynamics. MICS emulates a range of synaptic behaviors including associative learning. We further demonstrate its application in reservoir computing by performing handwritten digit classification and tic-tac-toe game and explore the system parameters that improve the computational performance. This droplet-based biomimetic synapse offers a potentially scalable and energy-efficient platform for next-generation neuromorphic computing systems.
Droplet interface bilayer↗
Development and evaluation of a predictive algorithm for telerobotic task complexity
There is a wide range of complexity in the various telerobotic servicing tasks performed in subsea, space, and hazardous material handling environments. Experience with telerobotic servicing has evolved into a knowledge base used to design tasks to be 'telerobot friendly.' This knowledge base generally resides in a small group of people. Written documentation and requirements are limited in conveying this knowledge base to serviceable equipment designers and are subject to misinterpretation. A mathematical model of task complexity based on measurable task parameters and telerobot performance characteristics would be a valuable tool to designers and operational planners. Oceaneering Space Systems and TRW have performed an independent research and development project to develop such a tool for telerobotic orbital replacement unit (ORU) exchange. This algorithm was developed to predict an ORU exchange degree of difficulty rating (based on the Cooper-Harper rating used to assess piloted operations). It is based on measurable parameters of the ORU, attachment receptacle and quantifiable telerobotic performance characteristics (e.g., link length, joint ranges, positional accuracy, tool lengths, number of cameras, and locations). The resulting algorithm can be used to predict task complexity as the ORU parameters, receptacle parameters, and telerobotic characteristics are varied.
Gernhardt, M. L.↗
Software Process Assurance for Complex Electronics
Complex Electronics (CE) now perform tasks that were previously handled in software, such as communication protocols. Many methods used to develop software bare a close resemblance to CE development. Field Programmable Gate Arrays (FPGAs) can have over a million logic gates while system-on-chip (SOC) devices can combine a microprocessor, input and output channels, and sometimes an FPGA for programmability. With this increased intricacy, the possibility of software-like bugs such as incorrect design, logic, and unexpected interactions within the logic is great. With CE devices obscuring the hardware/software boundary, we propose that mature software methodologies may be utilized with slight modifications in the development of these devices. Software Process Assurance for Complex Electronics (SPACE) is a research project that used standardized S/W Assurance/Engineering practices to provide an assurance framework for development activities. Tools such as checklists, best practices and techniques were used to detect missing requirements and bugs earlier in the development cycle creating a development process for CE that was more easily maintained, consistent and configurable based on the device used.
Plastow, Richard A.↗
Improving Trust in Deep Neural Networks with Nearest Neighbors
Deep neural networks are used increasingly for perception and decision-making in UAVs. For example, they can be used to recognize objects from images and decide what actions the vehicle should take. While deep neural networks can perform very well at complex tasks, their decisions may be unintuitive to a human operator. When a human disagrees with a neural network prediction, due to the black box nature of deep neural networks, it can be unclear whether the system knows something the human does not or whether the system is malfunctioning. This uncertainty is problematic when it comes to ensuring safety. As a result, it is important to develop technologies for explaining neural network decisions for trust and safety. This paper explores a modification to the deep neural network classification layer to produce both a predicted label and an explanation to support its prediction. Specifically, at test time, we replace the final output layer of the neural network classifier by a k-nearest neighbor classifier. The nearest neighbor classifier produces 1) a predicted label through voting and 2) the nearest neighbors involved in the prediction, which represent the most similar examples from the training dataset. Because prediction and explanation are derived from the same underlying process, this approach guarantees that the explanations are always relevant to the predictions. We demonstrate the approach on a convolutional neural network for a UAV image classification task. We perform experiments using a forest trail image dataset and show empirically that the hybrid classifier can produce intuitive explanations without loss of predictive performance compared to the original neural network. We also show how the approach can be used to help identify potential issues in the network and training process.
Lee, Ritchie↗
Acquisition and production of skilled behavior in dynamic decision-making tasks
This status report consists of a thesis entitled 'Ecological Task Analysis: A Method for Display Enhancements.' Previous use of various analysis processes for the purpose of display interface design or enhancement has run the risk of failing to improve user performance due to the analysis resulting in only a sequencial listing of user tasks. Adopting an ecological approach to performing the task analysis, however, may result in the necessary modeling of an unpredictable and variable task domain required to improve user performance. Kirlik has proposed an Ecological Task Analysis framework which is designed for this purpose. It is the purpose of this research to measure this framework's effectiveness at enhancing display interfaces in order to improve user performance. Following the proposed framework, an ecological task analysis of experienced users of a complex and dynamic laboratory task, Star Cruiser, was performed. Based on this analysis, display enhancements were proposed and implemented. An experiment was then conducted to compare this new version of Star Cruiser to the original. By measuring user performance at different tasks, it was determined that during early sessions, use of the enhanced display contributed to better user performance compared to that achieved using the original display. Furthermore, the results indicate that the enhancements proposed as a result of the ecological task analysis affected user performance differently depending on whether they are enhancements which aid in the selection of a possible action or in the performance of an action. Generalizations of these findings to larger, more complex systems were avoided since the analysis was only performed on this one particular system.
Kirlik, Alex↗
DS-GL: Advancing Graph Learning via Harnessing the Power of Nature within Dynamic Systems
With the rapid digitization of the world, an increasing number of real-world applications are turning to nonEuclidean data, modeled as graphs. Due to their intrinsic high complexity and irregularity, learning from graph data demands tremendous computational power. Recently, CMOS-compatible Ising machines, i.e., dynamic systems composed of CMOS components, have emerged as a new approach that harnesses the inherent power of natural annealing within dynamic systems to efficiently resolve binary optimization problems and have been adopted for traditional graph computation, such as max-cut. However, when performing complex Graph Learning (GL) tasks, Ising machines face significant hurdles: (i) they are inherently binary and thus ill-suited for real-valued problems; (ii) their expensive all-to-all coupling network that guarantees effective natural annealing poses daunting scalability concerns. To address these challenges, this paper proposes a nature-powered graph learning framework dubbed DS-GL, which is the first effort to transform the process of solving graph learning problems into the natural annealing process within a parameterized dynamic system embodied as a CMOS chip. To tackle the two major hurdles, DS-GL first augments the Ising machine architecture to modify the self-reaction term of its Hamiltonian function from linear to quadratic, effectively serving as an energy regulator. This adjustment maintains the system’s original physical interpretation while enabling it to process continuous, real-valued data. Second, to address the scaling issue, DS-GL further upgrades the real-valued dense Ising machine by decomposing it into a mesh-based multi-PE dynamic system that supports efficient distributed spatial-temporal co-annealing across different PEs through sparse interconnects. By exploiting the inherent sparsity and component structures in real-world graphs, DS-GL is able to map complex graph learning tasks onto the scalable dynamic system while maintaining high accuracy. Evaluations with three diverse GL applications across six real-world datasets, including traffic flow and COVID-19 prediction, show that DS-GL can deliver from 102× to 106× speedups and 500× energy reduction over Graph Neural Networks on GPUs, with 5% - 20% accuracy enhancement.
Song, Ruibing↗
Manual control using the matched manipulator control technique.
Manual control using matched manipulator technique, noting control task complexity decrease and improved performance
Herzog, J. H.↗
Operations Modeling for SSF Evolution
The operations required to support the on-orbit Space Station Freedom activities planned or being studied will be complex. Operational capability to perform tasks will be dependent on many factors such as manpower availability, logistics, other tasks being worked and Space Station configuration. This effort uses information available about these and other factors to perform operations analysis for given missions and determine the feasibility of target configuration concepts to support those missions. Studies have been conducted to determine processing requirements for a number of potential evolutionary missions on the Space Station Freedom. These studies have identified the need for growth of the Space Station in various ways. Some of the studies have dealt with the operational needs for the particular mission that they are concerned with, but none have looked at the total evolutionary operations requirement. In order to pursue the subject of overall on-orbit operations to any appreciable level of detail, data bases of operational on-orbit tasks need to be compiled, and an analysis tool is needed to assist the analyst. A number of existing operations tools have been reviewed, and none have been found to satisfactorily perform the functions needed to analyze integrated operations requirements for the evolutionary Space Station Freedom. However, during the tool review, some existing applications were found to provide subsets of the required functionality, and these are being considered for incorporation into the analysis tool.
Ganoe, George G.↗
ReQuBiS - Reconfigurable Quadrupedal-Bipedal Snake Robots
Self-assembling and self-reconfiguring robots have the ability to adapt to different or varying environments and carry out complex tasks. Robots competent of performing serpentine gaits and transforming into legged systems impart a wide range of locomotion schemes. Most of the reconfigurable robotic solutions consist of a large number of complicated mechanisms to detach and re-attach at different places making the system fragile and increase its size, weight and power (SWaP). We propose Reconfigurable Quadrupedal-Bipedal Snake Robots (ReQuBiS) to easily transform into these forms. Experimental results demonstrate the mobility in snake, quadruped and biped modes and transition between them with minimal change in modules.
Chiddarwar, Shital↗
Dissociation revisited - Workload and performance in a simulated flight task
The multiple resource model has been used as the theoretical basis for interpreting single-to-dual task changes in measures of performance and workload ratings. Inconsistent relationships among these measures have been termed dissociation. It is possible they are an artifact of the way performance and workload measures are collected; performance measures are available for the components of a complex task whereas workload ratings are integrated across all of the tasks performed within an interval of time. This study compared component ratings with global ratings and found that component ratings provide better information about subjects' task strategies and in interpreting the resultant relationships between workload ratings and performance.
King, Teresa↗
ChemGraph as an agentic framework for computational chemistry workflows
Atomistic simulations are essential in chemistry and materials science but remain challenging to run due to the expert knowledge required for the setup, execution, and validation stages of these calculations. We present ChemGraph, an agentic framework powered by artificial intelligence and state-of-the-art simulation tools to streamline and automate computational chemistry and materials science workflows. ChemGraph leverages graph neural network-based foundation models for accurate yet computationally efficient calculations and large language models (LLMs) for natural language understanding, task planning, and scientific reasoning to provide an intuitive and interactive interface. We evaluate ChemGraph across 13 benchmark tasks and demonstrate that smaller LLMs (GPT-4o-mini, Claude-3.5-haiku, Qwen-2.5-14B) perform well on simple workflows, while more complex tasks benefit from using larger models. Importantly, we show that decomposing complex tasks into smaller subtasks through a multi-agent framework enables GPT-4o to reach perfect accuracy and smaller LLMs to match or exceed single-agent GPT-4o's performance in these benchmarks.
Computational chemistry↗
A general architecture for intelligent training systems
A preliminary design of a general architecture for autonomous intelligent training systems was developed. The architecture integrates expert system technology with teaching/training methodologies to permit the production of systems suitable for use by NASA, other government agencies, industry, and academia in the training of personnel for the performance of complex, mission-critical tasks. The proposed architecture consists of five elements: a user interface, a domain expert, a training session manager, a trainee model, and a training scenario generator. The design of this architecture was guided and its efficacy tested through the development of a system for use by Mission Control Center Flight Dynamics Officers in training to perform Payload-Assist Module Deploys from the orbiter.
Loftin, R. Bowen↗
Exploration Medical Capability Clinical Decision Support System Concept of Operations
The Clinical Decision Support (CDS) project supports the Exploration Medical Capability (ExMC) Element of the Human Research Program (HRP). Specifically, the CDS project addresses the ExMC gap, Medical-701: Enhance medical capabilities within an exploration medical system. For long-duration, deep space missions, computational and data resources will play an important role in maintaining crew health, wellness and performance where the crew will need to be more self-reliant as we enter a new era in space exploration to return to the moon and explore Mars. These ambitious goals will require significant change in in-flight medical care due to constraints on mass, volume, power, crew time, skills reduction over time and medical evacuation capabilities. These constraints make it absolutely necessary to develop transformative solutions using new technologies. Unlike the current paradigm for crew health in low-Earth orbit missions that rely on constant communication with Mission Control, the deep space missions will experience communication delays and possibly, no communications for finite periods of time. Hence, crew health management will benefit from analytics’ capabilities to augment decision support. A comprehensive, multi-functional on-board clinical decision support system (CDSS) will help crews assess and diagnose conditions, decide appropriate responses, and guide the provision of tailored and evidence-based treatments, while reflecting contextual factors and constraints. The context may include present and historical data, viable diagnostic equipment, available supplies and medications, and vehicle and environmental health. Communication time with ground-based personnel is delayed or non-existent during significant portions of the mission so the crew will need to autonomously respond to health, performance and medical situations, particularly those that are unplanned. The CDSS must also provide additional capabilities as complex as training for an emergency situation while augmenting non-expert practitioner skillsets if the Crew Medical Officer (CMO) is incapacitated, and as routine as facilitating delayed communication with flight surgeons on the ground. The CDSS must connect complex issues involving health, wellness, task performance and environmental domains. Furthermore, CDSS functionality will focus on semi-autonomous and autonomous decision-making by the crew that is necessary to address challenges in executing a self-contained medical system that enables health care without assistance from ground clinical experts. The document, ExMC CDSS Architecture Recommendation, (HRP- 48032) establishes a description of the envisioned CDSS architecture. The analytics, descriptive or advanced, contained in a CDSS will interface with the integrated crew health and performance architecture that provides the appropriate data sets. The aim of the CDS project is to develop requirements for a CDSS through a series of test-bed prototype developments and demonstrations.
HRP↗
Importance of numerosity and distribution of articulations within the digits, wrists, and arms of telemanipulators confronted with dextrous assembly tasks
This study was conducted to evaluate the impact of successive exclusion of degrees-of-freedom within digits (thumb and index finger) and wrist when performing tasks demanding varying degrees of end effector kinematic complexity. The analysis was performed by successive constraining articulations in the fingers (thumb and index) and wrists or human subjects via splints, and then timing performance of part movement, positioning, and assembly operations requiring different levels of kinematic freedom. Analysis of performance times showed that transferring degrees-of-freedom, or articulations, from the digits to the wrist and arm had little or no material affect upon manipulation performance times if wrist and arm kinematics were unrestricted. If wrist and arm kinematics are not constrained and are of high-quality, our findings show that transferring kinematic freedom from the digits to the wrist and arms or remote end effectors offers limited or no consequence in terms of remote assembly operations. The importance of such findings, in terms of development and implementation of commercially attractive telemanipulators, is discussed.
Wiker, Steven F.↗
Performance-Aligned LLMs for Generating Fast HPC Code
Optimizing scientific software is a difficult task because codebases are often large and complex, and performance can depend upon several factors including the algorithm, its implementation, and hardware among others. Causes of poor performance can originate from disparate sources and be difficult to diagnose. Recent years have seen a multitude of work that use large language models (LLMs) to assist in software development tasks. However, these tools are trained to model the distribution of code as text, and are not specifically designed to understand performance aspects of code. In this work, we introduce a reinforcement learning based methodology to align the outputs of code LLMs with performance. This allows us to build upon the current code modeling capabilities of LLMs and extend them to generate better performing code. Here, we demonstrate that our fine-tuned model improves the expected speedup of generated code over base models for a set of benchmark tasks from 0.9 to 1.6 for serial code and 1.9 to 4.5 for OpenMP parallel code.
Computer science↗