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 253 records · Page 14

An Approach for Autonomy: A Collaborative Communication Framework for Multi-Agent Systems

Research done during the last three years has studied the emersion properties of Complex Adaptive Systems (CAS). The deployment of Artificial Intelligence (AI) techniques applied to remote Unmanned Aerial Vehicles has led the author to investigate applications of CAS within the field of Autonomous Multi-Agent Systems. The core objective of current research efforts is focused on the simplicity of Intelligent Agents (IA) and the modeling of these agents within complex systems. This research effort looks at the communication, interaction, and adaptability of multi-agents as applied to complex systems control. The embodiment concept applied to robotics has application possibilities within multi-agent frameworks. A new framework for agent awareness within a virtual 3D world concept is possible where the vehicle is composed of collaborative agents. This approach has many possibilities for applications to complex systems. This paper describes the development of an approach to apply this virtual framework to the NASA Goddard Space Flight Center (GSFC) tetrahedron structure developed under the Autonomous Nano Technology Swarm (ANTS) program and the Super Miniaturized Addressable Reconfigurable Technology (SMART) architecture program. These projects represent an innovative set of novel concepts deploying adaptable, self-organizing structures composed of many tetrahedrons. This technology is pushing current applied Agents Concepts to new levels of requirements and adaptability.

Dufrene, Warren Russell, Jr.↗

Developing Concepts of Operations Using Multi-Step Tool Techniques With Large Language Models

The National Aeronautics and Space Administration (NASA) Air Mobility Pathfinders (AMP) project is developing and evaluating concepts of operations (ConOps) for safe, secure, and scalable Urban Air Mobility (UAM) operations. The AMP project’s Operational Concepts, Architecture, and Requirements Integration (OCARI) Team is using a Model Based System Engineering (MBSE) approach for integration, interoperability, and traceability of Advanced Air Mobility (AAM) ecosystems centered around urban air taxi services. The team’s goal is to define structures and behaviors needed for system feasibility, readiness, and interoperability, establish a UAM knowledge base, and trace and validate assumptions and requirements relevant to AAM. NASA Langley Research Center (LaRC) is spearheading an innovative digital engineering approach to integrate, communicate, and facilitate the research of multi-modal transportation systems. The Knowledge-based Digital Platform (KbDP) is a concept being developed that ties the workflows of Project Managers (PM), Principal Investigators (PI), and System Engineers together across organizational boundaries. It does so through the management of an information database defined by mathematical, data science, and system engineering principles. Machine Learning (ML) algorithms play a key role in this concept by extracting meaningful knowledge from relational and graph databases, document repositories, and system artifacts, which the human user leverages to greatly improve the efficiency and effectiveness of their research. Recent advancements in the field of Large Language Models (LLMs), specifically models trained for tool use, such as Command-R , now allow for the reliable implementation of single-step and multi-step tool-centric systems. These techniques provide the LLM with a set of tools, in our case Python functions, that can be called on to answer a much wider range of questions compared to LLMs implemented using a traditional single-source or Retrieval Augmented Generation (RAG) approach. Through this method, the LLM can pull information from multiple data sources, such as relational or graph databases, document repositories, application programming interfaces (APIs), and SysML artifacts depending on the user’s question. The LLM can also output the information in a variety of different formats, using output generation tools, such as CSV, UML, or SysML artifacts. Additionally, tools can be assigned roles and can work together to provide answers to queries in an “agent” like approach, similar to that implemented by Microsoft’s AutoGen framework where different agents can converse with each other to accomplish tasks. Previously, our team developed a chatbot system with “agent like” functionality in the form of different “modes” the user could select from a user interface (UI), this architecture can be seen on the left in figure 1. Three different modes were implemented, the first mode allowed the LLM to utilize the structures and algorithms within a graph database to trace UAM requirements. The second mode gave the LLM access to a vector search capable of providing relevant information from thousands of document pages related to UAM ConOps and requirements. The third mode served as a general assistant where users could enter open-ended questions and custom prompts to utilize the LLM for different use-cases. This system improved the process surrounding generating and analyzing information related to UAM requirements, however, the implementation provided a clunky user experience. Users were required to know what mode to select within the UI in advance before entering their question to the selected tool. Moreover, the different tools were isolated from each other, they lacked bidirectional links that would allow for tools to collaborate to generate better responses. Our team is working on a new architecture, seen on the right in the below figure, with the goal to address many of the UX shortcomings of our original system while improving the accuracy and depth of responses from the LLM. This new system will automatically select the appropriate tool to use based off the user’s question. Each tool will be capable of calling on any of the other tools available to the LLM, resulting in a collaborative pipeline where tools can pass data between other tools until enough data is received to generate an answer to the user’s question. Using a locally deployed, open-source, LLM, the NASA OCARI team, in collaboration with Collins Aerospace, will implement a prototype application that will bridge knowledge across multiple sources to assist System Engineers (SEs) with requirements discovery and tracing, research question and use case identification, and assumption validation. Such a system will also allow SEs to more easily, and intuitively, explore the AAM ecosystem, ultimately improving the efficiency and effectiveness of the SE's research and decision-making processes surrounding ConOps development and validation. In this session, our team will provide a video demonstration of our new prototype architecture in action. We will also present an overview of our prototype system architecture and talk about its advantages over traditional LLM deployments along with how those advantages can provide additional value to the field of System Engineering.

systems engineering↗

Laser Induced Chemical Liquid Phase Deposition (LCLD)

Laser induced chemical deposition (LCLD) of metals onto different substrates attracts growing attention during the last decade. Deposition of metals onto the surface of dielectrics and semiconductors with help of laser beam allows the creation of conducting metal of very complex architecture even in 3D. In the processes examined the deposition occurs from solutions containing metal ions and reducing agents. The deposition happens in the region of surface irradiated by laser beam (micro reactors). Physics -chemical reactions driven by laser beam will be discussed for different metal-substrate systems. The electrical, optical, mechanical properties of created interfaces will be demonstrated also including some practical-industrial applications.

36 MATERIALS SCIENCE↗

CRAB3: Establishing a new generation of services for distributed analysis at CMS

In CMS Computing the highest priorities for analysis tools are the improvement of the end users ability to produce and publish reliable samples and analysis results as well as a transition to a sustainable development and operations model. To achieve these goals CMS decided to incorporate analysis processing into the same framework as data and simulation processing. This strategy foresees that all workload tools (TierO, Tier1, production, analysis) share a common core with long term maintainability as well as the standardization of the operator interfaces. The re-engineered analysis workload manager, called CRAB3, makes use of newer technologies, such as RESTFul based web services and NoSQL Databases, aiming to increase the scalability and reliability of the system. As opposed to CRAB2, in CRAB3 all work is centrally injected and managed in a global queue. A pool of agents, which can be geographically distributed, consumes work from the central services serving the user tasks. The new architecture of CRAB substantially changes the deployment model and operations activities. In this paper we present the implementation of CRAB3, emphasizing how the new architecture improves the workflow automation and simplifies maintainability. In particular, we will highlight the impact of the new design on daily operations.

Cinquilli, M.↗

Secure API-Driven Research Automation to Accelerate Scientific Discovery

The Secure Scientific Service Mesh (S3M) provides API-driven infrastructure to accelerate scientific discovery through automated research workflows. By integrating near real-time streaming capabilities, intelligent workflow orchestration, and fine-grained authorization within a service mesh architecture, S3M enables secure and flexible programmatic access to high performance computing (HPC) resources. This framework allows intelligent agents and experimental facilities to dynamically provision resources and execute complex workflows, accelerating experimental lifecycles, and enabling AI-augmented autonomous science. S3M establishes a modern foundation for scientific computing infrastructure that significantly reduces traditional barriers between researchers, computational resources, and experimental facilities.

Skluzacek, Tyler [ORNL] (ORCID:0000000322424931)↗

The AI Bus architecture for distributed knowledge-based systems

The AI Bus architecture is layered, distributed object oriented framework developed to support the requirements of advanced technology programs for an order of magnitude improvement in software costs. The consequent need for highly autonomous computer systems, adaptable to new technology advances over a long lifespan, led to the design of an open architecture and toolbox for building large scale, robust, production quality systems. The AI Bus accommodates a mix of knowledge based and conventional components, running on heterogeneous, distributed real world and testbed environment. The concepts and design is described of the AI Bus architecture and its current implementation status as a Unix C++ library or reusable objects. Each high level semiautonomous agent process consists of a number of knowledge sources together with interagent communication mechanisms based on shared blackboards and message passing acquaintances. Standard interfaces and protocols are followed for combining and validating subsystems. Dynamic probes or demons provide an event driven means for providing active objects with shared access to resources, and each other, while not violating their security.

Schultz, Roger D.↗

Virtualization - A Key Cost Saver in NASA Multi-Mission Ground System Architecture

With science team budgets being slashed, and a lack of adequate facilities for science payload teams to operate their instruments, there is a strong need for innovative new ground systems that are able to provide necessary levels of capability processing power, system availability and redundancy while maintaining a small footprint in terms of physical space, power utilization and cooling.The ground system architecture being presented is based off of heritage from several other projects currently in development or operations at Goddard, but was designed and built specifically to meet the needs of the Science and Planetary Operations Control Center (SPOCC) as a low-cost payload command, control, planning and analysis operations center. However, this SPOCC architecture was designed to be generic enough to be re-used partially or in whole by other labs and missions (since its inception that has already happened in several cases!)The SPOCC architecture leverages a highly available VMware-based virtualization cluster with shared SAS Direct-Attached Storage (DAS) to provide an extremely high-performing, low-power-utilization and small-footprint compute environment that provides Virtual Machine resources shared among the various tenant missions in the SPOCC. The storage is also expandable, allowing future missions to chain up to 7 additional 2U chassis of storage at an extremely competitive cost if they require additional archive or virtual machine storage space.The software architecture provides a fully-redundant GMSEC-based message bus architecture based on the ActiveMQ middleware to track all health and safety status within the SPOCC ground system. All virtual machines utilize the GMSEC system agents to report system host health over the GMSEC bus, and spacecraft payload health is monitored using the Hammers Integrated Test and Operations System (ITOS) Galaxy Telemetry and Command (TC) system, which performs near-real-time limit checking and data processing on the downlinked data stream and injects messages into the GMSEC bus that are monitored to automatically page the on-call operator or Systems Administrator (SA) when an off-nominal condition is detected. This architecture, like the LTSP thin clients, are shared across all tenant missions.Other required IT security controls are implemented at the ground system level, including physical access controls, logical system-level authentication authorization management, auditing and reporting, network management and a NIST 800-53 FISMA-Moderate IT Security plan Risk Assessment Contingency Plan, helping multiple missions share the cost of compliance with agency-mandated directives.The SPOCC architecture provides science payload control centers and backup mission operations centers with a cost-effective, standardized approach to virtualizing and monitoring resources that were traditionally multiple racks full of physical machines. The increased agility in deploying new virtual systems and thin client workstations can provide significant savings in personnel costs for maintaining the ground system. The cost savings in procurement, power, rack footprint and cooling as well as the shared multi-mission design greatly reduces upfront cost for missions moving into the facility. Overall, the authors hope that this architecture will become a model for how future NASA operations centers are constructed!

Ground System Architecture↗

Scalable/Secure Cooperative Algorithms and Framework for Extremely-high Penetration Solar Integration (SolarExPert) (Final Technical Report)

This SolarExPert project has developed a Sustainable Grid Platform (SGP) with scalable architecture of distributed control and optimization. The SGP consists of the following major functions: 1) an advanced grid architecture with hierarchical and distributed communication and control, combined with the OpenFMB standard and implemented on the Multi-Agent OpenDSS (MA-OpenDSS) platform; 2) an online distributed stochastic optimal power flow; 3) an online distributed system state estimation algorithm; 4) the distributed Volt/VAR optimization and frequency control algorithms; 5) the distributed distribution system restoration strategy. The developed SGP together with advanced functions are tested in 1 million (1M)-node distribution system on the MA-OpenDSS platform. Furthermore, the models and algorithms are tested in 100,000-node system HiL simulation, and also in P-HiL implementation with 100 physical devices. The developed functions haven been validated and tested on the selected actual distribution feeder with the data collected from the field of Maui Meadows in Hawaii. The distributed PV hosting capacities with cooperative Volt/VAR and Volt/VAR/Watt control are estimated and compared to provide recommendations for customers and the utility company.

14 SOLAR ENERGY↗

Model-based Executive Control through Reactive Planning for Autonomous Rovers

This paper reports on the design and implementation of a real-time executive for a mobile rover that uses a model-based, declarative approach. The control system is based on the Intelligent Distributed Execution Architecture (IDEA), an approach to planning and execution that provides a unified representational and computational framework for an autonomous agent. The basic hypothesis of IDEA is that a large control system can be structured as a collection of interacting agents, each with the same fundamental structure. We show that planning and real-time response are compatible if the executive minimizes the size of the planning problem. We detail the implementation of this approach on an exploration rover (Gromit an RWI ATRV Junior at NASA Ames) presenting different IDEA controllers of the same domain and comparing them with more classical approaches. We demonstrate that the approach is scalable to complex coordination of functional modules needed for autonomous navigation and exploration.

Finzi, Alberto↗

Serious Gaming for Test & Evaluation of Clean-Slate (Ab Initio) National Airspace System (NAS) Designs

Incremental approaches to air transportation system development inherit current architectural constraints, which, in turn, place hard bounds on system capacity, efficiency of performance, and complexity. To enable airspace operations of the future, a clean-slate (ab initio) airspace design(s) must be considered. This ab initio National Airspace System (NAS) must be capable of accommodating increased traffic density, a broader diversity of aircraft, and on-demand mobility. System and subsystem designs should scale to accommodate the inevitable demand for airspace services that include large numbers of autonomous Unmanned Aerial Vehicles and a paradigm shift in general aviation (e.g., personal air vehicles) in addition to more traditional aerial vehicles such as commercial jetliners and weather balloons. The complex and adaptive nature of ab initio designs for the future NAS requires new approaches to validation, adding a significant physical experimentation component to analytical and simulation tools. In addition to software modeling and simulation, the ability to exercise system solutions in a flight environment will be an essential aspect of validation. The NASA Langley Research Center (LaRC) Autonomy Incubator seeks to develop a flight simulation infrastructure for ab initio modeling and simulation that assumes no specific NAS architecture and models vehicle-to-vehicle behavior to examine interactions and emergent behaviors among hundreds of intelligent aerial agents exhibiting collaborative, cooperative, coordinative, selfish, and malicious behaviors. The air transportation system of the future will be a complex adaptive system (CAS) characterized by complex and sometimes unpredictable (or unpredicted) behaviors that result from temporal and spatial interactions among large numbers of participants. A CAS not only evolves with a changing environment and adapts to it, it is closely coupled to all systems that constitute the environment. Thus, the ecosystem that contains the system and other systems evolves with the CAS as well. The effects of the emerging adaptation and co-evolution are difficult to capture with only combined mathematical and computational experimentation. Therefore, an ab initio flight simulation environment must accommodate individual vehicles, groups of self-organizing vehicles, and large-scale infrastructure behavior. Inspired by Massively Multiplayer Online Role Playing Games (MMORPG) and Serious Gaming, the proposed ab initio simulation environment is similar to online gaming environments in which player participants interact with each other, affect their environment, and expect the simulation to persist and change regardless of any individual player's active participation.

Allen, B. Danette↗

Three dimensional nitrogen-doped graphene aerogels functionalized with melamine for multifunctional applications in supercapacitors and adsorption

In present work, we demonstrate an efficient and facile strategy to fabricate three-dimensional (3D) nitrogen-doped graphene aerogels (NGAs) based on melamine, which serves as reducing and functionalizing agent of graphene oxide (GO) in an aqueous medium with ammonia. Benefiting from well-defined and cross-linked 3D porous network architectures, the supercapacitor based on the NGAs exhibited a high specific capacitance of 170.5 F g{sup −1} at 0.2 A g{sup −1}, and this capacitance also showed good electrochemical stability and a high degree of reversibility in the repetitive charge/discharge cycling test. More interestingly, the prepared NGAs further exhibited high adsorption capacities and high recycling performance toward several metal ions such as Pb{sup 2+}, Cu{sup 2+} and Cd{sup 2+}. Moreover, the hydrophobic carbonized nitrogen-doped graphene aerogels (CNGAs) showed outstanding adsorption and recycling performance for the removal of various oils and organic solvents. - Graphical abstract: Three-dimensional nitrogen-doped graphene aerogels were prepared by using melamine as reducing and functionalizing agent in an aqueous medium with ammonia, which showed multifunctional applications in supercapacitors and adsorption. - Highlights: • Three-dimensional nitrogen-doped graphene aerogels (NGAs) were prepared. • Melamine was used as reducing and functionalizing agent. • NGAs exhibited relatively good electrochemical properties in supercapacitor. • NGAs exhibited high adsorption performance toward several metal ions. • CNGAs showed outstanding adsorption capacities for various oils and solvents.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Software for Automation of Real-Time Agents, Version 2

Version 2 of Closed Loop Execution and Recovery (CLEaR) has been developed. CLEaR is an artificial intelligence computer program for use in planning and execution of actions of autonomous agents, including, for example, Deep Space Network (DSN) antenna ground stations, robotic exploratory ground vehicles (rovers), robotic aircraft (UAVs), and robotic spacecraft. CLEaR automates the generation and execution of command sequences, monitoring the sequence execution, and modifying the command sequence in response to execution deviations and failures as well as new goals for the agent to achieve. The development of CLEaR has focused on the unification of planning and execution to increase the ability of the autonomous agent to perform under tight resource and time constraints coupled with uncertainty in how much of resources and time will be required to perform a task. This unification is realized by extending the traditional three-tier robotic control architecture by increasing the interaction between the software components that perform deliberation and reactive functions. The increase in interaction reduces the need to replan, enables earlier detection of the need to replan, and enables replanning to occur before an agent enters a state of failure.

Fisher, Forest↗

Sonochemical fabrication of Cu{sub 2}O@C/graphene nanohybrid with a hierarchical architecture

A three-dimensional Cu{sub 2}O@C/graphene nanohybrid hierarchical sphere-like architecture has been successfully synthesized by a sonochemical method. The as-prepared nanocomposite has been characterized using various techniques including powder X-ray diffraction, scanning electron microscopy, field emission transmission electron microscopy, thermogravimetric analysis, X-ray photoelectron spectroscopy and Fourier transmission infrared spectrometry. Secondary big hierarchical Cu{sub 2}O@C spheres (diameter of ∼200 nm) consist of primary small Cu{sub 2}O@C spheres (diameter of 5–10 nm), which are uniformly distributed on the graphene surface. In the primary small Cu{sub 2}O@C sphere, amorphous carbon is formed by the oxidation–reduction reaction between 1-aminopyrene and copper(II) acetate in the sonochemical process. Our facile sonochemical method may apply to the preparation of other metal oxide/graphene hierarchical nanohybrids. - Highlights: • The Cu{sub 2}O@C/graphene nanohybrid hierarchical architecture was obtained. • Secondary big hierarchical Cu{sub 2}O@C spheres (diameter of ∼200 nm) consist of primary small Cu{sub 2}O@C spheres (diameter of 5–10 nm). • 1-Aminopyrene serves as a “binding agent” operating through the π–π interactions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Impact Studies and Cooperative Voltage Control for High PV Penetration

The drastically increasing penetration of photovoltaic (PV) has posed tremendous challenges in operation, control, and protection in distribution networks. The over-voltage issue is one of major concerns for extremely high PV penetration (e.g. 100%). It is critical to identify the most sensitive locations to high PV penetration and apply effective methods to mitigate the over-voltage issue. This paper introduced a novel greedy search algorithm to locate worst-case scenarios of over-voltage in both distributed and centralized PV configurations, as well as a cooperative voltage control method to mitigate over-voltage. The search algorithm has the computational advantage over existing stochastic approach and provides more insights in impact studies. The cooperative control method relies upon reactive power compensation to converge system voltage to a predefined value within standard limits. Moreover, a clustering strategy is applied to divide the network into several clusters, within which all nodes share information with neighboring nodes following a scalable communication architecture. The control algorithms and communication hierarchies presented in this paper have been implemented into OpenDSS to form the Multi Agent (MA) OpenDSS package, which has been made available to the reader. The effectiveness of developed methods are demonstrated through simulation studies on the IEEE 123-node test feeder under four different load levels.

Cooperative control↗

Power-Law Entanglement and Hilbert Space Fragmentation in Nonreciprocal Quantum Circuits

Quantum circuits utilizing measurement to evolve a quantum wave function offer a new and rich playground to engineer unconventional entanglement dynamics. Here, in this work, we introduce a hybrid, nonreciprocal setup featuring a quantum circuit, whose updates are conditioned on the state of a classical dynamical agent. In our example the circuit is represented by a Majorana quantum chain controlled by a classical N-state Potts chain undergoing pair flips. The local orientation of the classical spins controls whether randomly drawn local measurements on the quantum chain are allowed or not. This imposes a dynamical kinetic constraint on the entanglement growth, described by the transfer matrix of an N-colored loop model. It yields an equivalent description of the circuit by an SU(N)-symmetric Temperley-Lieb Hamiltonian or by a kinetically constrained surface growth model for an N-component height field. For N = 2, we find a diffusive growth of the half-chain entanglement toward a stationary profile S(L) ~ L 1/2 for L sites. For N ≥ 3, the kinetic constraints impose Hilbert space fragmentation, yielding subdiffusive growth toward S(L) ~ L 0.57 . This showcases how the control by a classical dynamical agent can enrich the entanglement dynamics in quantum circuits, paving a route toward novel entanglement dynamics in nonreciprocal hybrid circuit architectures.

1-dimensional spin chains↗

Accurate Consensus-based Distributed Averaging with Variable Time Delay in Support of Distributed Secondary Control Algorithms

We report that distributed secondary control has been widely used in hierarchical control structures, where multiple distributed generators (DGs) need to coordinate to regulate system voltage and frequency. In these systems, consensus algorithms determine the average of a group of dynamic states (e.g. voltages measured by a group of DGs). To be useful, consensus algorithms must be computationally efficient, stable and accurate. In practice, numerous practical implementation challenges significantly affect the consensus equilibrium. In this paper, we quantify the accuracy deviations of the distributed average observer algorithms proposed in the literature to demonstrate the problems with the state-of-the-art distributed averaging techniques. A novel approach is proposed that achieves accurate average tracking in the presence of time-varying communication delays among agents. In our implementation, time synchronization of all distributed controllers is enabled by a novel software platform, called Resilient Information Architecture Platform for the Smart Grid (RIAPS). The proposed distributed average observer is implemented on hardware controllers and its effectiveness is validated in a controller hardware-in-the-loop testbed.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Cross-Domain Reasoning for Neuromorphic Model Design

Designing performant neuromorphic models requires reasoning across neuroscience, neuromorphic computing, and machine learning, making it a natural target for cross-domain hypothesis generation. Our primary contribution is a multi-corpus knowledge graph spanning all three domains, which we show substantially increases cross-domain retrieval novelty over single-corpus baselines. We additionally introduce NeuKReAct, an agentic reasoning framework that iteratively retrieves from this graph and synthesizes design hypotheses via a step-by-step blackboard architecture, enabling structured compartmentalization of design decisions. Lastly, we introduce an execution head that translates hypotheses into structured design documents and runnable code. We evaluate novelty using a combinatorial creativity metric that measures cross-domain retrieval distance across the citation graph. Our results confirm that corpus breadth is the dominant driver of novelty. Moreover, we highlight a concrete instance of the novelty-utility tradeoff within NeuKReAct, underscoring a need for joint creativity evaluation, balancing both novelty and utility.

Ramavarapu, Vikram [ORNL] (ORCID:0009000188757213)↗

Bootstrapping Multi-Agent Unmanned Aerial Vehicle (UAV) System Integration Using Ground-Based Assets: Lessons Learned

In support of the Autonomy Teaming and TRAjectories for Complex Trusted Operational Reliability (ATTRACTOR) project, a fleet of unmanned ground vehicles (UGVs) was developed as a test and evaluation (T\&E) platform to reduce system integration gaps between simulation and live flight hardware. While simulation and hardware-in-the-loop bench testing provide adequate environments for preliminary validation, differences in system deployment architecture, software interfaces, and hardware infrastructure increase the risks to safety, property, and the project. Given ATTRACTOR’s goal of establishing a basis of certification of trust and trustworthiness in multi-agent autonomous systems, bridging these gaps was critical to successful project execution and feasibility assessment. In this paper we present the UGV fleet and its role in speeding up system integration, smoothing the transition from simulation to flight, and providing researchers an easy-to-use hardware test bed. An overview of the hardware and software on-board the vehicles is provided along with supporting infrastructure. The system integration process is documented including results in supporting both the overarching design reference mission (DRM) of ATTRACTOR and individual research efforts conducted since the creation of the fleet. Finally, we discuss the practical lessons learned regarding the testing, deployment, and operation of multi-agent autonomous systems.

Matthew P. Vaughan↗