Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “agent architectures”

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 217 records · Page 12

A Study on Efficient Reinforcement Learning Through Knowledge Transfer

Although Reinforcement Learning (RL) algorithms have made impressive progress in learning complex tasks over the past years, there are still prevailing short-comings and challenges. Specifically, the sample-inefficiency and limited adaptation across tasks often make classic RL techniques impractical for real-world applications despite the gained representational power when combining deep neural networks with RL, known as Deep Reinforcement Learning (DRL). Recently, a number of approaches to address those issues have emerged. Many of those solutions are based on smart DRL architectures that enhance single task algorithms with the capability to share knowledge between agents and across tasks by introducing Transfer Learning (TL) capabilities. Here this survey addresses strategies of knowledge transfer from simple parameter sharing to privacy preserving federated learning and aims at providing a general overview of the field of TL in the DRL domain, establishes a classification framework, and briefly describes representative works in the area.

97 MATHEMATICS AND COMPUTING↗

From Rules to Reasoning: A Survey of Large Language Model-Based Approaches to Scientific Hypothesis and Idea Generation

Scientific hypothesis generation represents a fundamental challenge in contemporary research due to exponentially expanding literature volumes and increasing disciplinary specialization. Large language models (LLMs) have emerged as transformative tools for automated scientific discovery, moving beyond traditional rule-based and literature-mining approaches. Four paradigmatic approaches define current LLM-driven hypothesis generation: direct prompting and fine-tuning methods, knowledge-enhanced frameworks integrating retrieval-augmented generation (RAG), multi-agent collaborative systems simulating research teams, and reasoning-focused approaches implementing cognitive architectures. Domain-specific applications demonstrate statistical equivalence to human expert performance in social psychology, experimental validation in biomedical research, and near-expert quality in astronomy. Evaluation methodologies encompass human expert assessment, LLM-as-judge frameworks, and comprehensive benchmarking systems. Technical challenges include hallucination management, knowledge integration limitations, and balancing novelty with feasibility. Future directions emphasize hybrid neural-symbolic architectures and sophisticated human-AI collaboration models for responsible scientific discovery acceleration.

AI-driven discovery↗

Prototyping a large-scale distributed system for the Great Observatories era - NASA Astrophysics Data System (ADS)

The NASA Astrophysics Data System (ADS) is a distributed information system intended to support research in the Great Observatories era, to simplify access to data, and to enable simultaneous analyses of multispectral data sets. Here, the user agent and interface, its functions, and system components are examined, and the system architecture and infrastructure is addressed. The present status of the system and related future activities are examined.

Shames, Peter↗

An Architecture for Controlling Multiple Robots

The Control Architecture for Multirobot Outpost (CAMPOUT) is a distributed-control architecture for coordinating the activities of multiple robots. In the CAMPOUT, multiple-agent activities and sensor-based controls are derived as group compositions and involve coordination of more basic controllers denoted, for present purposes, as behaviors. The CAMPOUT provides basic mechanistic concepts for representation and execution of distributed group activities. One considers a network of nodes that comprise behaviors (self-contained controllers) augmented with hyper-links, which are used to exchange information between the nodes to achieve coordinated activities. Group behavior is guided by a scripted plan, which encodes a conditional sequence of single-agent activities. Thus, higher-level functionality is composed by coordination of more basic behaviors under the downward task decomposition of a multi-agent planner

Aghazarian, Hrand↗

Home Energy Management System - VOLTTRON Integration

In most Home Energy Management Systems (HEMS) available in the market, different devices running different communication protocols cannot interact with each other and exchange information. As a result of this integration, the information about different devices running different communication protocol can be accessible by other agents and devices running on VOLTTRON platform. The integration process can be used by any HEMS available in the market regardless of the programming language they use. If the existing HEMS provides an Application Programming Interface (API) based on the RESTFul architecture, that API can be used for integration. Our candidate HEMS in this project is home-assistant (Hass). An agent is implemented which can communicate with the Hass API and receives information about the devices loaded on the API. The agent publishes the information it receives on the VOLTTRON message bus so other agents can have access to this information. On the other side, for each type of devices, an agent is implemented such as Climate Agent, Lock Agent, Switch Agent, Light Agent, etc. Each of these agents is subscribed to the messages published on the message bus about their associated devices. These agents can also change the status of the devices by sending appropriate service calls to the API. Other agents and services on the platform can also access this information and coordinate their decision-making process based on this information.

Zandi, Helia↗

Home Energy Management System - VOLTTRON Integration

In most Home Energy Management Systems (HEMS) available in the market, different devices running different communication protocols cannot interact with each other and exchange information. As a result of this integration, the information about different devices running different communication protocol can be accessible by other agents and devices running on VOLTTRON platform. The integration process can be used by any HEMS available in the market regardless of the programming language they use. If the existing HEMS provides an Application Programming Interface (API) based on the RESTFul architecture, that API can be used for integration. Our candidate HEMS in this project is home-assistant (Hass). An agent is implemented which can communicate with the Hass API and receives information about the devices loaded on the API. The agent publishes the information it receives on the VOLTTRON message bus so other agents can have access to this information. On the other side, for each type of devices, an agent is implemented such as Climate Agent, Lock Agent, Switch Agent, Light Agent, etc. Each of these agents is subscribed to the messages published on the message bus about their associated devices. These agents can also change the status of the devices by sending appropriate service calls to the API. Other agents and services on the platform can also access this information and coordinate their decision-making process based on this information.

Zandi, Helia↗

Building Energy Management Open Source Software

This is the repository for Building Energy Management Open Source Software (BEMOSS), which is an open source operating system that is engineered to improve sensing and control of equipment in small- and medium-sized commercial buildings. BEMOSS offers the following key features: (1) Open source, open architecture – BEMOSS is an open source operating system that is built upon VOLTTRON – a distributed agent platform developed by Pacific Northwest National Laboratory (PNNL). BEMOSS was designed to make it easy for hardware manufacturers to seamlessly interface their devices with BEMOSS. Software developers can also contribute to adding additional BEMOSS functionalities and applications. (2) Plug & play – BEMOSS was designed to automatically discover supported load controllers (including smart thermostats, VAV/RTUs, lighting load controllers and plug load controllers) in commercial buildings. (3) Interoperability – BEMOSS was designed to work with load control devices form different manufacturers that operate on different communication technologies and data exchange protocols. (4) Cost effectiveness – Implementation of BEMOSS deemed to be cost-effective as it was built upon a robust open source platform that can operate on a low-cost single-board computer, such as Odroid. This feature could contribute to its rapid deployment in small- or medium-sized commercial buildings. (5) Scalability and ease of deployment – With its multi-node architecture, BEMOSS provides a distributed architecture where load controllers in a multi-floor and high occupancy building could be monitored and controlled by multiple single-board computers hosting BEMOSS. This makes it possible for a building engineer to deploy BEMOSS in one zone of a building, be comfortable with its operation, and later on expand the deployment to the entire building to make it more energy efficient. (6) Ability to provide local and remote monitoring – BEMOSS provides both local and remote monitoring ability with role-based access control. (7) Security – In addition to built-in security features provided by VOLTTRON, BEMOSS provides enhanced security features, including BEMOSS discovery approval process, encrypted core-to-node communication, thermostat anti-tampering feature and many more. (8) Support from the Advisory Committee – BEMOSS was developed in consultation with an advisory committee from the beginning of the project. BEMOSS advisory committee comprises representatives from 22 organizations from government and industry.

Source record↗

Building Energy Management Open Source Software

This is the repository for Building Energy Management Open Source Software (BEMOSS), which is an open source operating system that is engineered to improve sensing and control of equipment in small- and medium-sized commercial buildings. BEMOSS offers the following key features: (1) Open source, open architecture – BEMOSS is an open source operating system that is built upon VOLTTRON – a distributed agent platform developed by Pacific Northwest National Laboratory (PNNL). BEMOSS was designed to make it easy for hardware manufacturers to seamlessly interface their devices with BEMOSS. Software developers can also contribute to adding additional BEMOSS functionalities and applications. (2) Plug & play – BEMOSS was designed to automatically discover supported load controllers (including smart thermostats, VAV/RTUs, lighting load controllers and plug load controllers) in commercial buildings. (3) Interoperability – BEMOSS was designed to work with load control devices form different manufacturers that operate on different communication technologies and data exchange protocols. (4) Cost effectiveness – Implementation of BEMOSS deemed to be cost-effective as it was built upon a robust open source platform that can operate on a low-cost single-board computer, such as Odroid. This feature could contribute to its rapid deployment in small- or medium-sized commercial buildings. (5) Scalability and ease of deployment – With its multi-node architecture, BEMOSS provides a distributed architecture where load controllers in a multi-floor and high occupancy building could be monitored and controlled by multiple single-board computers hosting BEMOSS. This makes it possible for a building engineer to deploy BEMOSS in one zone of a building, be comfortable with its operation, and later on expand the deployment to the entire building to make it more energy efficient. (6) Ability to provide local and remote monitoring – BEMOSS provides both local and remote monitoring ability with role-based access control. (7) Security – In addition to built-in security features provided by VOLTTRON, BEMOSS provides enhanced security features, including BEMOSS discovery approval process, encrypted core-to-node communication, thermostat anti-tampering feature and many more. (8) Support from the Advisory Committee – BEMOSS was developed in consultation with an advisory committee from the beginning of the project. BEMOSS advisory committee comprises representatives from 22 organizations from government and industry.

Source record↗

Redox regulation of Ran GTPase

Ran, a small Ras-like GTP-binding nuclear protein, plays a key role in modulation of various cellular signaling events including the cell cycle. This study shows that a cellular redox agent (nitrogen dioxide) facilitates Ran guanine nucleotide dissociation, and identifies a unique Ran redox architecture involved in that process. Sequence analysis suggests that Dexras1 and Rhes GTPases also possess the Ran redox architecture. As Ran releases an intact nucleotide, the redox regulation mechanism of Ran is likely to differ from the radical-based guanine nucleotide modification mechanism suggested for Ras and Rho GTPases. These results provide a mechanistic reason for the previously observed oxidative stress-induced perturbation of the Ran-mediated nuclear import, and suggest that oxidative stress could be a factor in the regulation of cell signal transduction pathways associated with Ran.

60 APPLIED LIFE SCIENCES↗

Atomic structure of a toxic, oligomeric segment of SOD1 linked to amyotrophic lateral sclerosis (ALS)

Fibrils and oligomers are the aggregated protein agents of neuronal dysfunction in ALS diseases. Whereas we now know much about fibril architecture, atomic structures of disease-related oligomers have eluded determination. Here, we determine the corkscrew-like structure of a cytotoxic segment of superoxide dismutase 1 (SOD1) in its oligomeric state. Mutations that prevent formation of this structure eliminate cytotoxicity of the segment in isolation as well as cytotoxicity of the ALS-linked mutants of SOD1 in primary motor neurons and in a Danio rerio (zebrafish) model of ALS. Cytotoxicity assays suggest that toxicity is a property of soluble oligomers, and not large insoluble aggregates. Here, our work adds to evidence that the toxic oligomeric entities in protein aggregation diseases contain antiparallel, out-of-register β-sheet structures and identifies a target for structure-based therapeutics in ALS.

59 BASIC BIOLOGICAL SCIENCES↗

Future applications of artificial intelligence to Mission Control Centers

Future applications of artificial intelligence to Mission Control Centers are presented in the form of the viewgraphs. The following subject areas are covered: basic objectives of the NASA-wide AI program; inhouse research program; constraint-based scheduling; learning and performance improvement for scheduling; GEMPLAN multi-agent planner; planning, scheduling, and control; Bayesian learning; efficient learning algorithms; ICARUS (an integrated architecture for learning); design knowledge acquisition and retention; computer-integrated documentation; and some speculation on future applications.

Friedland, Peter↗

Agentic AI vs ML-Based Autotuning: A Comparative Study for Loop Reordering Optimization

High Performance Computing (HPC) applications rely heavily on code optimizations to achieve good performance on modern CPU and GPU architectures. Traditional Machine Learning auto-tuning approaches have demonstrated success in exploring high-dimensional spaces, but they often require expensive compile-run evaluations and lack adaptability for large HPC applications. The recent advances in Large Language Models (LLMs) and Agentic AI systems raise intriguing questions about the potential of these approaches to address specific optimization methodologies. This work aims to answer an essential question for the HPC community: “How Agentic AI Systems Compare to Traditional ML Autotuning Techniques?” To address this question, we present a comparative analysis between a traditional ML-based optimization approach and an Agentic AI system, evaluating their respective capabilities and limitations for loop-level optimization. In addition, we introduced a new Agentic AI system named LoopGen-AI using three different Large Language Models: GPT-4.1, Claude 4.0, and Gemini 2.5. A key finding is that LoopGen-AI achieves competitive per-formance with only a few program runs, the reasoning logs from the agents revealed that their decisions rely heavily on the combination of semantic understanding of the target kernel with dynamic feedback from the environment, highlighting a promising new dimension in performance tuning. In contrast, ML-based autotuners focus on statistical exploration, and require orders of magnitude more runs to reach peak performance. Additionally, our analysis shows that prompt engineering, particularly using Persona + Context Manager patterns, significantly impacts the effectiveness of Agentic AI. Our results indicate that while Agentic AI systems are not yet a complete replacement for ML-based autotuners, it can effectively complement traditional methods.

Rosas, Miguel Romero↗

Autonomy Architectures for a Constellation of Spacecraft

Until the past few years, missions typically involved fairly large expensive spacecraft. Such missions have primarily favored using older proven technologies over more recently developed ones, and humans controlled spacecraft by manually generating detailed command sequences with low-level tools and then transmitting the sequences for subsequent execution on a spacecraft controller. This approach toward controlling a spacecraft has worked spectacularly on previous missions, but it has limitations deriving from communications restrictions - scheduling time to communicate with a particular spacecraft involves competing with other projects due to the limited number of deep space network antennae. This implies that a spacecraft can spend a long time just waiting whenever a command sequence fails. This is one reason why the New Millennium program has an objective to migrate parts of mission control tasks onboard a spacecraft to reduce wait time by making spacecraft more robust. The migrated software is called a "remote agent" and has 4 components: a mission manager to generate the high level goals, a planner/scheduler to turn goals into activities while reasoning about future expected situations, an executive/diagnostics engine to initiate and maintain activities while interpreting sensed events by reasoning about past and present situations, and a conventional real-time subsystem to interface with the spacecraft to implement an activity's primitive actions. In addition to needing remote planning and execution for isolated spacecraft, a trend toward multiple-spacecraft missions points to the need for remote distributed planning and execution. The past few years have seen missions with growing numbers of probes. Pathfinder has its rover (Sojourner), Cassini has its lander (Huygens), and the New Millenium Deep Space 3 (DS3) proposal involves a constellation of 3 spacecraft for interferometric mapping. This trend is expected to continue to progressively larger fleets. For example, one mission proposed to succeed DS3 would have 18 spacecraft flying in formation in order to detect earth-sized planets orbiting other stars. A proposed magnetospheric constellation would involve 5 to 500 spacecraft in Earth orbit to measure global phenomena within the magnetosphere. This work describes and compares three autonomy architectures for a system that continuously plans to control a fleet of spacecraft using collective mission goals instead of goals or command sequences for each spacecraft. A fleet of self-commanding spacecraft would autonomously coordinate itself to satisfy high level science and engineering goals in a changing partially-understood environment making feasible the operation of tens or even a hundred spacecraft (such as for interferometry or plasma physics missions). The easiest way to adapt autonomous spacecraft research to controlling constellations involves treating the constellation as a single spacecraft. Here one spacecraft directly controls the others as if they were connected. The controlling "master" spacecraft performs all autonomy reasoning, and the slaves only have real-time subsystems to execute the master's commands and transmit local telemetry/observations. The executive/diagnostics module starts actions and the master's real-time subsystem controls the action either locally or remotely through a slave. While the master/slave approach benefits from conceptual simplicity, it relies on an assumption that the master spacecraft's executive can continuously monitor the slaves' real-time subsystems, and this relies on high-bandwidth highly-reliable communications. Since unintended results occur fairly rarely, one way to relax the bandwidth requirements involves only monitoring unexpected events in spacecraft. Unfortunately, this disables the ability to monitor for unexpected events between spacecraft and leads to a host of coordination problems among the slaves. Also, failures in the communications system can result in losing slaves. The other two architectures improve robustness while reducing communications by progressively distributing more of the other three remote agent components across the constellation. In a teamwork architecture, all spacecraft have executives and real-time subsystems - only the leader has the planner/scheduler and mission manager. Finally, distributing all remote agent components leads to a peer-to-peer approach toward constellation control.

Barrett, Anthony↗

Variational quantum reinforcement learning via evolutionary optimization

Abstract Recent advances in classical reinforcement learning (RL) and quantum computation point to a promising direction for performing RL on a quantum computer. However, potential applications in quantum RL are limited by the number of qubits available in modern quantum devices. Here, we present two frameworks for deep quantum RL tasks using gradient-free evolutionary optimization. First, we apply the amplitude encoding scheme to the Cart-Pole problem, where we demonstrate the quantum advantage of parameter saving using amplitude encoding. Second, we propose a hybrid framework where the quantum RL agents are equipped with a hybrid tensor network-variational quantum circuit (TN-VQC) architecture to handle inputs of dimensions exceeding the number of qubits. This allows us to perform quantum RL in the MiniGrid environment with 147-dimensional inputs. The hybrid TN-VQC architecture provides a natural way to perform efficient compression of the input dimension, enabling further quantum RL applications on noisy intermediate-scale quantum devices.

97 MATHEMATICS AND COMPUTING↗

Wintap

Wintap is an extensible, general purpose agent framework for the Windows operating system. It provides an easy-to-use plugin architecture, an integrated streaming analytics engine, and real-time event subscriptions for host-based data. Wintap can support a wide variety of applications across production operations, cyber security operations, and cyber security research.

Frye, DavidJ↗

Cognitive Model of Trust Dynamics Predicts Human Behavior within and between Two Games of Strategic Interaction with Computerized Confederate Agents

When playing games of strategic interaction, such as iterated Prisoner’s Dilemma and iterated Chicken Game, people exhibit specific within-game learning (e.g., learning a game’s optimal outcome) as well as transfer of learning between games (e.g., a game’s optimal outcome occurring at a higher proportion when played after another game). The reciprocal trust players develop during the first game is thought to mediate transfer of learning effects. Recently, a computational cognitive model using a novel trust mechanism has been shown to account for human behavior in both games, including the transfer between games. We present the results of a study in which we evaluate the model’s a priori predictions of human learning and transfer in 16 different conditions. The model’s predictive validity is compared against five model variants that lacked a trust mechanism. The results suggest that a trust mechanism is necessary to explain human behavior across multiple conditions, even when a human plays against a non-human agent. The addition of a trust mechanism to the other learning mechanisms within the cognitive architecture, such as sequence learning, instance-based learning, and utility learning, leads to better prediction of the empirical data. It is argued that computational cognitive modeling is a useful tool for studying trust development, calibration, and repair.

59 BASIC BIOLOGICAL SCIENCES↗

Integrating planning, execution, and learning

To achieve the goal of building an autonomous agent, the usually disjoint capabilities of planning, execution, and learning must be used together. An architecture, called MAX, within which cognitive capabilities can be purposefully and intelligently integrated is described. The architecture supports the codification of capabilities as explicit knowledge that can be reasoned about. In addition, specific problem solving, learning, and integration knowledge is developed.

Kuokka, Daniel R.↗