TECHNICAL TRACKS: ABSTRACTS
WEDNESDAY, APRIL 29 (10:30 AM – 12:30 PM)
TRACK 1 Future Digital Processing including AI Techniques (Hudgins)
Receiver-Demodulators and the Winds of Change
Anad Kelkar, CDSI
As digital antenna systems proliferate, their capabilities have expanded exponentially. This is evident at the Pt. Mugu Sea Range (PMSR), where current systems simultaneously track over 50 data sources. However, conventional Receiver/Demodulator (RX/DEMOD) hardware increasingly dominates the system’s Size, Weight, and Power (SWaP) footprint. While manageable for fixed sites, transportable systems require a departure from the “one receiver per stream” model. CDSI is collaborating with manufacturers to develop digital-format solutions that transport data streams to RX/DEMOD functions entirely free of analog components. This presentation explores the trade-space and developments driving this transition
Standards-Based Digital RF Demodulation
Alex Ortiz, Parraid ** Gold Sponsor
Digital antenna systems capable of simultaneously tracking multiple data sources are increasingly deployed across test ranges. While these systems natively produce digital RF samples, they are commonly adapted to support current generation analog RF and IF telemetry receivers, limiting the benefits inherent to digital antenna systems. This presentation describes a modular, standards based, multi channel digital demodulator designed to interface directly with digital RF samples and perform telemetry demodulation entirely in the digital domain, while providing Telemetry over IP (TMoIP) outputs. The architecture leverages VITA 49 Digital IF Interoperability (DIFI) over SFP28 for digital RF sample input and IRIG 218 TMoIP for output, providing an interoperable, vendor neutral path to digital telemetry range modernization.
MIDAPS – Capability realized through Bi-directional TM
Mark Wigent, Laulima Systems LLC
The advent of bi-directional telemetry brings significant new capabilities that will transform the way tests can be conducted. The Mission Information Distribution and Processing System (MIDAPS) is a system for acquiring, processing, distributing, and analyzing telemetry data that has been designed from the ground up to deliver new capabilities in bi-directional telemetry. This presentation will provide an overview of MIDAPS capabilities.
NASA’s Quantum Sensor with Artificial NV Diamond (QSAND) for Magnetic Anomaly Navigation (MAGNAV)
Mark Hagiwara, NASA Armstrong
Modern navigation systems rely heavily on GPS. GPS signals are vulnerable to jamming and spoofing. Aircraft may experience GPS signal degradation or complete loss due to environmental factors. NASA’s Quantum Sensor with Artificial NV Diamond (QSAND) project aims to solve this problem using Magnetic Anomaly Navigation (MagNav), which uses Earth’s crustal magnetic anomalies as a reference for positioning. Quantum technologies harness quantum mechanics to accelerate advances in computing, communication, and sensing. Quantum sensing uses the unique properties of quantum systems to measure physical quantities, including time, inertia, magnetic fields and electric fields. This discussion firstly gives an overview of quantum sensing along with its advantages, potential impact and practical challenges. The discussion then outlines QSAND’s goals and presents an overview of past and present project development. It covers navigation in GPS-denied environments using an NV center quantum sensor. The presentation then details previous demonstrations of scalar and vector magnetic field sensing as well as the development of a compact prototype system. Then, the discussion outlines improvements in system performance and a future project approach for technology maturation. Finally, an overview of other potential applications using NV center diamonds is presented.
TRACK 2 Phased Array Tech-Applied to Current Test Requirements – (Young)
CRYPTO STAYS HOME: Black Data on Demand
Mike Robson, Raven Defense** Networking Sponsor
Here we discuss how Raven Defense has conducted real-time relay of full-frame, BLACK encrypted data from a deployed their Maritime Networked Test Asset (MaNTA) platform without hosting sensitive cryptographic assets onboard. The enabling technology is Starshield, due to its low latency data transport capability in Low Earth orbit LEO. While it does have a high Maximum Information Rate (MIR) for uplink, these high speeds are burst speeds and not sustained above 5 Mbps; inadequate for critical high data rate telemetry streams.
By leveraging multiple bonded Starshield terminals, the system aggregates uplink capacity to reliably achieve sustained data rates of 20 Mbps, and we have demonstrated the ability to transmit up to 40 Mbps, albeit with a lot of dropouts. This capability allows high-volume telemetry streams to be transmitted in real-time from the vessel to a government owned downlink terminal and onto a secure network for processing, where encryption, key management, and data analysis occur within a controlled and hardened environment. The approach reduces the operational and security burden on the forward-deployed platform while maintaining end-to-end data integrity and mission assurance.
The bonded-link architecture introduces resilience and performance gains by distributing traffic across multiple (N+1) terminals, accounting for data overhead costs and Virtual Private Network (VPN) consumption. Advanced traffic shaping, packetization strategies, and link management ensure stable throughput for full-frame telemetry, even under contested or variable conditions.
AV-Multi-band, Multi-target Digital Antenna TM Tracker
Adam Hartman, AeroVironment (AV)
AeroVironment (AV) in partnership with Georgia Tech Research Institute and the Test Resource Management Center has advanced the state of airborne telemetry through the collaborative development and deployment of long-range radio frequency (RF) phased array telemetry antennas, tailored for the U.S. Government’s SkyRange program. Designed for integration with both the RQ-4 Global Hawk and MQ-9 Reaper platforms, these arrays enable reliable, over-the-horizon telemetry capture from multiple simultaneous data streams. Leveraging a shared architectural framework, subsystem commonality, and standardized building blocks, AV’s RF phased array solution meets the rigorous environmental challenges posed by extreme altitudes and temperature variations, and incorporates robust, ruggedized electronics. Both the RQ-4 and MQ-9 benefit from a fully digital system design that enhances interference management, streamlines future mission augmentation, and facilitates capability upgrades without requiring hardware replacement. This modular, future-looking approach ensures superior performance, high reliability, and operational flexibility to meet evolving test and evaluation requirements across the SkyRange initiative. The presentation will detail system development, deployment outcomes, and the transformative advantages provided by AV’s phased array technology for next-generation telemetry operations.
Modular Antenna Systems for Airborne Telemetry Applications
Brian Mulford, CDSI
Reliable telemetry systems are essential for supporting weapons test operations and validating weapons system performance. Recognizing the need for more flexible and expansive test range capabilities, the concept of an airborne test range—now known as SkyRange—was developed to extend telemetry coverage and support advanced weapons testing with a quick-response solution. Creative Digital Systems Integration (CDSI) was tasked with developing a dynamic telemetry solution capable of operating aboard the Global Hawk aircraft platform. Since 2021, CDSI’s Universal Beamforming Technology (UBT) systems have successfully supported hypersonic weapons testing by reliably collecting and transmitting telemetry data from flight test vehicles worldwide. Building upon this proven capability, a more robust and advanced version of the UBT system has been developed and is currently being integrated into additional Global Hawk aircraft as well as MQ-9 Reaper pod configurations. This evolution in airborne telemetry technology represents a significant advancement in test range capability and is increasingly shaping the future of weapons test infrastructure.
Multi-band Antenna Tiles for Fixed & Transportable Ground Stations
Anand Kelkar, CDSI
Leveraging our legacy product, the Digital Beamforming Module: DBM(v2), CDSI introduced the DBM(v4), featuring direct RF sampling and hybrid FPGA-driven processing. The system’s flexible fabric allows for seamless switching between wideband (L+S or C-band) and narrower band-specific configurations via firmware updates. We will discuss the performance and design trade-space relevant to deploying these modules in fixed and transportable ground stations.
TRACK 3 AI -Powering Revolutionary Change to DoW Systems –M. Cook
Atari-Enabling AI at the Edge
James Charleston Jr., USAF USMC 812th AITS/ENIE
With need to accelerate Test and Evaluation Flight Test aircraft need to incorporate Artificial Intelligence and Machine Learning (AI/ML) processing capability in physically constrained environments. The AI Test Aircraft Research and Integration (ATARI) platform incorporates edge computing to run government developed software and AI/ML models on test aircraft. ATARI leverages containerization for modularity and ease of integration, which enables users to deploy custom solutions efficiently to support a variety of test scenarios. The platform features an Apache Kafka data backbone, ensuring near real-time, low-loss access to test data, to accelerate going from raw data to useful information. The first project to be implemented using ATARI is an on-aircraft speech-to-text voice transcription system, utilizing OpenAI’s Whisper model. ATARI represents a transformative approach to integrating AI/ML into flight test instrumentation, by combining advanced computational capabilities with robust data handling. DISTRIBUTION STATEMENT A. Approved for public release; Distribution is unlimited 412TW-PA- 26083
Beyond the Noise: AI-Powered Threat Detection
Jason Schalow, USAF, 412th Comms Squadron
Cybersecurity analysts are confronted with the monumental task of monitoring vast quantities of network authentication data to identify malicious activities. The sheer volume of this traffic makes it exceedingly difficult to detect sophisticated cyber threats, which are often designed to blend in with normal network behavior. Manual analysis is no longer sufficient to secure these complex environments. To address this critical challenge, the 412 TW’s RDT&E Enterprise Defensive Cyber Operations Response Element (REDCORE) is developing a novel, AI-driven deep learning framework designed for advanced anomaly detection. This system learns the normal patterns and rhythms of network activity, including the typical sequences of events. By creating a dynamic “blueprint” of what constitutes normal behavior, the model can accurately identify subtle deviations that may indicate a security threat. When tested against standard cybersecurity benchmarks, the model proved highly effective. It successfully identified threats with greater accuracy and a lower rate of false alarms compared to other established methods. Future work will focus on integrating this powerful detection engine with a large language model to create a comprehensive AI assistant. This will empower security analysts by providing them with deeper insights and significantly enhancing their ability to defend against the evolving threat landscape.
The Agentic Whitebox: A Novel Framework for Verifiable and Auditable AI-Assisted Test & Evaluation
Kevin Kerkvliet, ATEC
This paper introduces the Agentic Whitebox Test & Evaluation (AWT&E) framework, a novel methodology for integrating artificial intelligence into the Department of Defense’s Test & Evaluation (T&E) lifecycle. The framework directly confronts the critical challenge of “Black Box” AI, where systems produce answers without auditable reasoning, making them unsuitable for high-stakes evaluation. The AWT&E proposes a federated system of specialized AI agents, designated “Daxton,” designed to augment human test officers, not replace them. The system automates the laborious processes of data correlation, test plan validation, and results analysis while mandating that every significant conclusion is accompanied by a mathematically sound, verifiable proof of its reasoning. This approach shifts the T&E paradigm from slow, manual analysis to rapid, AI-assisted, evidence-based evaluation that is both transparent and trustworthy. Methodology: The core of the AWT&E methodology is the “Proof of Reasoning,” a principle requiring every AI-generated insight to be justified through a human-relatable, scientific metric. This is achieved through a suite of specialized metrics designed for different phases of the T&E process. For data link analysis, the “Interface Reliability Score (IRS)” uses statistical process control to quantify network stability in terms of standard deviations, translating Z-scores into intuitive “Green/Yellow/Red” operational impact alerts. For test planning, the “Test Plan Coverage & Sufficiency Score (TCSS)” applies a quantitative framework to measure the depth and breadth of requirements coverage. For results analysis, the “Finding Confidence Index (FCI)” employs a multi-source corroboration algorithm, weighting evidence from disparate, authoritative data sources (e.g., system logs, Test Incident Reports, observer notes) to produce a probabilistic confidence score for every proposed finding. This methodology ensures that rather than receiving a simple answer, the human evaluator receives the full, auditable “work” behind the conclusion, enabling true Human-on-the-Loop governance. The entire framework is designed to be built upon existing authoritative data sources and government cloud infrastructure, such as the Army’s cArmy and Enterprise LLM Workspace. Expected Contribution: The expected contribution of this work is threefold. First, it provides a practical, implementable framework that moves the concept of “Responsible AI” from policy into practice within the T&E domain. Second, it offers the T&E community a validated set of “White Box” metrics (IRS, TCSS, FCI) that can be adopted to scientifically measure the performance of AI-enabled systems and the validity of AI-generated conclusions. Finally, it presents a clear, data-driven path to dramatically accelerating the T&E lifecycle, reducing report generation from months to days. By solving the explainability problem, the AWT&E framework enables the DoD to leverage the speed of AI without sacrificing the scientific rigor and accountability essential for delivering safe, effective, and reliable capabilities to the Warfighter.
AI-Enabled Systems for Military Test & Evaluation
Steve Seiden, Acquired Data Systems
The Department of Defense is increasingly evaluating artificial intelligence as an enabling capability within complex weapon, vehicle, and sustainment systems. In the field of vehicle maintenance, most technicians must perform manual visual inspection to identify and diagnose instances of wear, corrosion, or damage. Manually identifying issues requires both time and experience on the part of the maintainer, and diagnostic errors or repair delays may affect equipment availability and readiness. Augmenting historically manually performed inspections with AI-based computer vision represents a means to enhance the accuracy, reliability, and repeatability of vehicle wear analysis and damage detection, and enables maintainers to increase their productivity by reducing the time required to identify and diagnose issues. This abstract presents a technical approach for incorporating AI-enabled perception systems into military T&E environments, using an AI-based computer vision application for ground vehicle tread wear analysis as an illustrative example. While the system discussed in this presentation relates to track pad analysis, the AI-based computer vision approach may be employed to augment virtually any visual assessment-based maintenance testing process, such as wear analysis on powertrain components, corrosion detection, or damage assessments.
This presentation will discuss the repeatable framework for developing and evaluating AI-enabled computer vision systems within military T&E pipelines. By treating AI capabilities as measurable, instrumented components of a test architecture, the approach supports defensible assessment of autonomy-adjacent technologies, AI-assisted maintenance, and future human-machine teaming concepts. This framework provides a pathway for transitioning AI from experimental capability to testable, certifiable military systems.
TRACKS WEDNESDAY, APRIL 29 (1:30PM – 3:30 PM)
TRACK 4 Phased Array: Applications for Improved Capabilities (Young)
ARMCHAIR OPERATORS: Operating BLOS Telemetry Missions from Home
Mike Robson, Raven Defense
Here we explore a paradigm shift in Beyond-Line-Of-Sight (BLOS) flight test support enabled by low-latency satellite communications (SATCOM). Historically, such over the horizon telemetry collection has been constrained to manned vessels outfitted with locally operated sensors and equipment. Depending on deployed capabilities, the mission crew (guests, or non-boat crew) regularly number >>10 souls, while the boat crew will add another 11 or more for a standard Offshore Service Vessel (OSV), and large research vessel will sail with around 50. Shore to shore durations number 3 to 5 weeks, and frequently go over holidays and family events, equating to high monetary and human costs.
In an effort to alleviate these costs, Raven Defense has invested heavily in the ability to transition to distributed environments where operators can execute and monitor mission systems remotely. Leveraging emerging commercial broadband constellations—most notably low Earth orbit (LEO) systems such as Starlink/Starshield—these architectures provide the bandwidth, latency, and reliability required to support real-time telemetry, sensor operations, Flight Termination System (FTS) command and control, and situational awareness from virtually any location. This evolution enables the use of unmanned or minimally crewed platforms to collect critical test data with fewer resources and less cost.
The concept of “armchair operators” introduces real opportunities for modern test enterprises. It enhances flexibility, reduces personnel footprint in hazardous environments, and allows subject matter experts to participate in missions without geographic constraints.
How Phased Array TM Antennas are Supporting Expansion of the Sea Ranges
Adam Hartman, AeroVironment (AV)
AeroVironment (AV) has pioneered the development and deployment of advanced long-range RF phased array telemetry antennas to enhance the U.S. Navy’s test and evaluation capabilities at the Point Mugu Sea Range. These state-of-the-art sensors, installed on San Nicolas Island, offer multi-stream, multi-directional telemetry signal reception across the L, S, and C frequency bands, capturing data from hundreds of nautical miles away. This technology represents a major leap forward in the Navy’s ability to add complexity and scale to test events, significantly augmenting the range’s operational reach. Developed in close collaboration with the U.S. Navy, the Test Resource Management Center, Georgia Tech Research Institute, and BAE Systems, AV’s phased array solution introduces unmatched flexibility and resilience. The fully digital system architecture provides advanced interference management, facilitates rapid mission adaptation, and enables seamless future upgrades—all without requiring hardware modifications. This presentation will outline the design, development process, and operational deployment of the antennas, and will highlight the strategic benefits delivered to the Navy’s testing infrastructure by this collaborative, next-generation telemetry capability.
THE GHOST CRUISE: RAPTR’S Solo Ocean Crossing
Mike Robson, Raven Defense
Here we discuss the ground-breaking operation of our unmanned, remotely operated/ autonomous surface vessel, Wraith, equipped with the Raven Defense Advance Phased Array Telemetry Resource (RAPTR). In February of 2026, we self-funded an Atlantic deployment demonstration in support of to support the Multi-Service Advanced Capability Hypersonics Test Bed (MACH-TB) hypersonic flight testing of HASTE and DART. Wraith was launched from the Virginia coast and autonomously transited hundreds of miles offshore where the platform endured harsh maritime conditions, including the largest North Atlantic winter storm in a decade, while maintaining system integrity and mission readiness. By eliminating the need for crewed ships, we demonstrated that this approach significantly reduces operational cost, risk to personnel, and logistical complexity, while enabling persistent, precisely positioned telemetry collection in remote ocean test ranges.
Wraith successfully transited to the Test Support Position (TSP) providing an additional telemetry node in the test architecture, where it captured high-rate data from early flight out of NASA’s Wallops Flight Facility (WFF) through the Second Engine Cut Off (SECO) event of the Cassowary test flight mission. The integration of autonomous navigation, resilient power systems, and stabilized phased array tracking enables reliable performance despite dynamic sea states and environmental uncertainty.
Phased Array Antennas Spanning Multiple UAV Platforms Empower SkyRange TM Gathering Operations
Adam Hartman, AeroVironment (AV)
AeroVironment (AV) in partnership with Georgia Tech Research Institute and the Test Resource Management Center has advanced the state of airborne telemetry through the collaborative development and deployment of long-range radio frequency (RF) phased array telemetry antennas, tailored for the U.S. Government’s SkyRange program. Designed for integration with both the RQ-4 Global Hawk and MQ-9 Reaper platforms, these arrays enable reliable, over-the-horizon telemetry capture from multiple simultaneous data streams. Leveraging a shared architectural framework, subsystem commonality, and standardized building blocks, AV’s RF phased array solution meets the rigorous environmental challenges posed by extreme altitudes and temperature variations, and incorporates robust, ruggedized electronics.Both the RQ-4 and MQ-9 benefit from a fully digital system design that enhances interference management, streamlines future mission augmentation, and facilitates capability upgrades without requiring hardware replacement. This modular, future-looking approach ensures superior performance, high reliability, and operational flexibility to meet evolving test and evaluation requirements across the SkyRange initiative. The presentation will detail system development, deployment outcomes, and the transformative advantages provided by AV’s phased array technology for next-generation telemetry operations.
TRACK 5 Telemetry Systems Breakthrough for the Test Environment (Elder)
X-59 Instrumentation System Overview
Matthew Versteeg, NASA Armstrong
The X-59 QueSST aircraft represents a critical step toward quiet supersonic flight. Central to its success is a sophisticated instrumentation system designed to capture high-fidelity aerodynamic, structural, and systems data in support of NASA’s Low-Boom Flight Demonstration program. This presentation provides an overview of the X-59 instrumentation architecture and the associated data acquisition and telemetry systems. Particular emphasis will be placed on the integration challenges encountered during development and the practical solutions implemented to overcome them — ranging from hardware retrofits and software tools to real-time data visualization and rapid troubleshooting capabilities. Attendees will gain insight into how these instrumentation solutions enable efficient testing, reduce operational dependencies, and maintain schedule performance in a complex flight research environment.
Spectrum Sharing Technologies for 5G Coexistence
Rob Zeiglar, Peraton Labs
Coming Soon
New Concepts for 5G Range Implementation
Achilles Kogiantis, Altio Labs
As the 3GPP 5G standard evolves to address complex new use cases, the commercial-off-the-shelf (COTS) hardware ecosystem is maturing in tandem. Testing ranges, having established baseline operational experience with 5G, are now positioned to push the technology into more demanding environments.This presentation cross-references the latest 5G standards roadmap with the demands of modern test ranges—including extreme propagation distances, high-velocity links, remote deployment constraints, low-latency needs and high reliability. By mapping the gap between current COTS capabilities and future standards, we identify immediate and near-term opportunities to evolve the 5G testing range infrastructure.
High efficiency Bi-Directional Amplifier for High PAPR 5G Wave Forms
Achilles Kogiantis, Altio Labs
5G networks utilize OFDM as the primary modulation method due to its advantages in long-range wireless communication, while using high-order modulation and multi-carrier waveforms to achieve very high data rates. As a result, these waveforms have very high peak-to-average power ratios that make amplification difficult and inefficient. This high Peak-to-Average Power Ratio (PAPR) greatly impacts the efficiency of RF power amplifiers needed for extended links since power amplifiers operate most efficiently near their saturation point. However, driving them into saturation results in peak signal clipping, leading to out of band interference. To minimize clipping and comply with spectral mask regulations that prevent interference, conventional amplifiers are typically operated with power backoff, which significantly lowers efficiency. In this talk we discuss our approach to linearizing the amplifier by a combination of crest factor reduction and digital pre-distortion (DPD) techniques. We demonstrate the substantial improvements we can achieve with CFR/DPD in terms of adjacent channel leakage ratio (ACLR), Error Vector Magnitude (EVM), and protection of IRIG-106 receivers. We discuss the development plan for a self-contained module to support a high efficiency bi-directional amplifier for 5G Aeronautical Mobile Telemetry (AMT) applications.
TRACK 6 Topics supporting T&E (Alich)
Taming the Beast: Secure Retrieval-Augmented Generation for T&E Report Automation in Classified Environments
Michael Soltys-Kulinicz, Ph.D, GBL Systems Corp
Test and evaluation (T&E) organizations face a persistent productivity bottleneck: analysts spend significant time manually searching through TEMPs, test plans, MIL-STDs, range SOPs, and historical reports to produce compliant deliverables. This manual process is slow, error-prone, and difficult to scale — directly conflicting with the imperative to Test for Less.Retrieval-Augmented Generation (RAG) offers a practical, deployable solution. RAG connects a Large Language Model (LLM) to an organization’s actual document corpus, enabling it to retrieve relevant content and synthesize accurate, source-cited answers in natural language. When combined with few-shot prompting — providing the model with example report templates — RAG can accelerate the generation of test reports, evaluation summaries, and after-action reviews while maintaining format compliance and traceability to source documents.This presentation draws on production experience deploying RAG systems for the U.S. Navy at NSWC Port Hueneme Division (PHD). The authors built and fielded two operational systems: one that captures domain expert knowledge as a queryable knowledge base, and another that automates contract document processing. Both systems run in AWS GovCloud and use Amazon Bedrock for LLM inference and text embeddings, backed by a Milvus vector database and the Haystack RAG framework. This proven technology stack is now being adapted for the Intelligent Range Optimization Scheduler (IROS) at NAWCWD to provide natural language access to range scheduling policies, DoDI 5000 series test policy, and facility documentation.A critical challenge for DoD adoption of RAG is deploying AI systems that handle Controlled Unclassified Information (CUI) and classified data in compliance with NIST 800-171, CMMC, and DFARS 252.204-7012. This presentation addresses the security architecture required for production RAG in sensitive environments, using AWS Bedrock in GovCloud as a concrete example. Key security considerations include: accessing Bedrock through VPC endpoints (AWS PrivateLink) to keep all inference traffic off the public internet; encrypting data at rest with customer-managed KMS keys rather than AWS-owned keys to maintain auditable key management under NIST 800-171 controls 3.13.10 and 3.3.1; enforcing TLS 1.2+ for all data in transit per control 3.13.8; and implementing role-based access controls and audit logging across the RAG pipeline. The presentation also discusses the trade-offs between commercial LLM services and self-hosted open-source models for air-gapped and disconnected environments, drawing on direct comparative testing.Attendees will leave with a clear understanding of: (1) how RAG works and why it outperforms general-purpose LLMs for domain-specific T&E tasks; (2) the concrete architecture and technology stack behind a production DoD RAG deployment; (3) the security controls required to operate RAG systems with CUI and classified data; and (4) how few-shot prompting transforms RAG from a question-answering tool into an automated report generation capability that reduces analyst workload and accelerates T&E deliverables.Keywords: Retrieval-Augmented Generation, artificial intelligence, test and evaluation, report automation, AWS Bedrock, CUI compliance, NIST 800-171, GovCloud, few-shot prompting, open-source AI
T-34 Autonomy & Flight Research Testbed TAFT
Jinu Idicula, NASA Armstong
NASA Armstrong is developing the T-34 aircraft as a flexible flight research capability. The T-34 Autonomy & Flight Research Testbed (TAFT) seeks to address the persistent challenges in evaluating autonomy, sensors, and other new ideas. An external pod with a reconfigurable layout, software & experiment development kit and modular open architecture tailored to researcher needs will help in overcoming these barriers.
This capability will be developed in three phases. In the first phase, a research flight computer, experimental software, and sensors will provide data display and advisory information to the backseat pilot through a tablet interface. The second phase will expand the testbed to integrate aircraft system data, flight computer software, and external wing-mounted sensor data through a common data bus, supporting more sophisticated experimentation and real-time feedback. The final phase will add a mechanical cable-puller system with limited control authority linked to a research flight computer capable of directly commanding the aircraft. This allows the pilot to engage the autonomy or experiment while being able to override or disengage the system at any time. Once fully complete, TAFT will provide researchers with an “autonomy sandbox” to rapidly test their ideas.
From a test and evaluation perspective, the platform is intended to support early and continuous experimentation for autonomy and other novel technologies. TAFT can incrementally evaluate performance, collect representative flight data, assess sensor behavior, and refine algorithms in an operationally relevant environment. Once fully operational, we believe TAFT will provide a framework for increasing flight test throughput & reducing the time required to evaluate autonomy, sensors, and other novel concepts
Avoiding Vendor Lock in – AI Procurement for DOW
Michael Soltys-Kulinicz, Ph.D, GBL Systems Corp
Vendor lock-in poses significant risks to Department of War (DoW) artificial intelligence (AI) procurement, threatening technological sovereignty, limiting competition, and increasing long-term costs. Lock-in occurs when proprietary architectures, data formats, or integration dependencies limit the government’s ability to switch suppliers, reduce costs, or adapt systems over time. This paper examines the unique challenges facing DoW AI acquisition and provides actionable strategies for maintaining platform independence. We analyze technical approaches including containerization, open standards adoption, and Infrastructure-as-Code, alongside procurement reforms such as modular contracting and performance-based acquisition. Case studies from the Chief Digital and AI Office, Army Enterprise Cloud, and Navy AI modernization programs demonstrate that vendor independence is achievable at operational scale without compromising mission effectiveness. We present a phased implementation roadmap for DoW organizations seeking to reduce vendor dependencies while maintaining security, reliability, and operational readiness. The strategies outlined address technical, commercial, operational, and strategic dimensions of vendor lock-in, providing comprehensive guidance for acquisition professionals and technical teams.
Leveraging Retrieval Augmented Generation (RAG) to Mitigate Data Security Risks in Generative AI Models
Jeff Kalibjian, Peraton Labs
Using sensitive information to train (e.g. From Scratch Training, FST or Fine-Tuned Training, FTT) generative AI models is problematic due to the lack of deterministic cybersecurity controls available in generative AI models which could protect such information from being disclosed to unauthorized third parties. Retrieval Augmented Generation (RAG) technology can allow a generative AI solution to access sensitive information, but prevent such information from being embedded in a generative AI model. After reviewing the present state of generative AI data security challenges, RAG technologies will be reviewed with emphasis on application to testing environments.
TRACKS THURSDAY, APRIL 30 (1:30PM – 3:30 PM)
TRACK 7 Leveraging 5G and Beyond for T&E (Aguirre)
From Spectrum Planning to Measurement: Building a Unified Analytics Portal for Test Range Spectrum Operations
Achilles Kogiantis, Altio Labs
The T&E community increasingly depends on RF spectrum–intensive systems, yet planning, coordination, and measurement of spectrum use across ranges remain fragmented across many independent tools and data sources. The Consolidated Spectrum Tools (CST) addresses this challenge by providing a unified, web-based portal that integrates spectrum planning data, scheduling information, and real-time measurement analytics into a single operational environment. Rather than focusing solely on coordination workflows, CST emphasizes the fusion of spectrum sensing data, mission scheduling inputs, and historical utilization metrics to create a continuously updated spectrum operational picture for test ranges. Through a modular, plugin-driven architecture, CST enables advanced analytics, interference detection, and utilization reporting while allowing ranges to tailor capabilities to their specific instrumentation and operational needs. The result is a scalable platform that allows spectrum managers, test planners, and engineers to move from static frequency coordination toward data-driven spectrum management supported by live measurements and analytics across the range enterprise.
Development for Flight Test Telemetry
Rob Zeigler, Peraton Labs
Coming Soon
5G Telemetry for the Holloman High Speed Test Track
Achilles Kogiantis, Altio Labs
The Holloman High Speed Test Track enables ground-based testing at supersonic and hypersonic speeds. As the telemetry needs are expanding to address the more complex modern systems under test, wideband telemetry systems are sought, with 5G being one of possible solutions. We present the challenges the test track testing needs pose for the telemetry link along with its topology and operational aspects. We describe how a 5G telemetry system can be configured, and what customizations will be needed to keep a radio link alive during a test. We construct a lab equivalent of the radio conditions of the high-speed test track and we attempt to confirm via emulation that the 5G setup can maintain a stable radio link. Finally, we describe additional aspects that constrain the possible telemetry solutions such as sled flexibility and environmentals.
ARL 5G Network Implementation and Test Results for UAS Testing
Achilles Kogiantis, Altio Labs
We detail the development, implementation, and initial testing of a fully mobile, 5G network deployed at the DEVCOM Army Research Laboratory’s Robotics Research Collaboration Campus (R2C2) at the Aberdeen Proving Ground. Designed to support unmanned aerial systems (UAS) and unmanned ground vehicles (UGV), the network was engineered to meet advanced connectivity requirements, facilitating real-time data visualization, bulk data offloading, and reliable communication between platforms and personnel. Key challenges addressed in this project include developing an uplink-heavy network architecture supporting a Massive MIMO Radio Access Network (RAN) to maximize capacity and uplink throughput. To support varied testing environments, the system is designed to integrat multiple 5G NR 3GPP Release 16 compliant User Equipment (UE) form factors, including ruggedized routers for UGVs and low-SWaP modules tailored for UAS applications. Network performance was initially modeled to predict coverage across the Graces Quarters peninsula for ground vehicles, as well as aerial coverage for UAS.We demonstrate the robust network capabilities with the tower fully deployed at 150ft high, with measured data rates significantly making use of MIMO techniques. We show uplink data rates supported in heavily wooded areas at various locations. The presentation concludes with an overview of optimizations we performed to fine tune the network system performance.
TRACK 8 Advancements in Traditional Telemetry Systems (Green)
Wireless Airborne Instrumentation Flight Test Results
Benjamin Baird, 896th Test Support Squadron, Eglin AFB
This presentation addresses the use of wireless technology in the area of airborne instrumentation with a focus on the reduction of aircraft downtime required for instrumentation installation. This application of wireless technology has the potential to significantly reduce the timely installation of wiring into and out of the cockpit and to other difficult to access locations throughout the aircraft. Recent flight test results and progress will be discussed.
Adapting Telemetry and Recording Systems to new Advanced Air Mobility Platforms and Unmanned Aerial Systems (UAS)
Dan Green, Safran DSI
Traditionally, flight test solutions were designed for flight test platforms, where space, cabling, and equipment size are not major constraints. Today, rapid technological advancement is transforming the aerospace landscape, and flight testing with it.Engineers now face recurring challenges: limited installation space, complex wiring, constrained power availability, and increasingly compact airframes. Our skies are no longer dominated solely by traditional fixed-wing aircraft and helicopters. They are increasingly populated by Advanced Air Mobility (AAM) platforms and Unmanned Aerial Systems (UAS), which introduce a fundamentally different set of design constraints.These next-generation platforms still require robust telemetry and instrumentation, but the industry must adapt its approach. AAM aircraft, particularly eVTOLs, are engineered to be compact, weight-optimized, aerodynamically efficient, and highly integrated. Every cubic inch and every pound directly impact battery range, payload, and overall performance. Oversized or heavy test equipment competes with critical systems, especially energy storage, and can significantly affect mission viability.As a result, flight test solutions must become more modular, scalable, and lightweight. The level of data precision and robustness remains critical, but architectures must now prioritize size, weight, and power efficiency. This requires rethinking system design, power management, and data acquisition and recording strategies to align with the operational realities of modern AAM and UAS platforms.In this evolving environment, virtualization-based recording architectures are becoming increasingly relevant. By decoupling data acquisition and recording functions from dedicated hardware and leveraging existing onboard computing resources, flight test engineers can significantly reduce system footprint, weight, and power consumption. Multi-format recording capabilities further enhance flexibility, enabling compatibility with established standards while preserving data integrity and traceability. When designed for seamless integration within the aircraft’s native systems, virtualized solutions simplify installation, reduce cabling complexity, and support more adaptable test configurations without compromising performance.This paper will present an overview of these evolving flight test strategies, highlighting virtualized approaches that maintain precision and reliability while addressing the unique constraints of AAM and UAS platforms.
The Use of TENA in a Network Centric Telemetry System
Mark Wigent, Laulima Systems LLC
This presentation will describe the use of the Test and Training Enabling Architecture (TENA) in the development of a network-centric, end-to-end data distribution and analysis system for acquiring, processing, and displaying dynamically selectable network-telemetered data
EPIC – Enhancing Parachutes by Instrumenting the Canopy; Flexible Strain Sensing, Development Approach and Test
Mark Hagiwara, NASA Armstrong
Parachutes are critical for bringing humans and valuable cargo to Earth and Mars surface. As Low Earth Orbit becomes the next major economic zone, and Mars becomes increasingly a target of interest, the need for parachute space-surface delivery is becoming rapidly more frequent while payloads are simultaneously trending larger and heavier. Previous tests on new parachute designs have sometimes shown unexpected and undesired results compared to current simulation models. A large part of the problem is today’s models are mostly unvalidated due to a lack of direct canopy measurements including localized strain measurements – this leads to an incomplete understanding of parachute dynamics and improper designs. NASA’s Enhancing Parachutes by Instrumenting the Canopy (EPIC) project aims to remedy the measurement problem by developing a sensor-instrumentation suite for localized canopy measurement for use in subsonic and eventually supersonic parachute deployment environments. Chief among the sensors is a flexible strain sensor capable of stretching with the canopy fabric while not impeding parachute performance and still producing reliable data. This discussion goes over EPIC’s goals and presents an overview of past and current project development approaches. The discussion covers a previous stiffness and adherence test which showed sensor operation without parachute interference. The discussion then covers previous bubble, dynamic, and lakebed drop tests, describes the data, and presents achievements so far. A full-scale parachute testing roadmap with a future project approach is then presented. Finally, the discussion presents possible applications beyond parachutes.