Capabilities - R4C Tech

Our Expertise Domains

Advanced cyber-physical intelligence software systems for defense and aerospace applications.

R4C Tech builds software that turns complex, multi-source operational data into reliable, decision-ready systems for defense and aerospace. Our expertise spans space domain awareness, distributed systems, signal and sensor processing, and applied machine learning, proven in production at the U.S. Space Force SDA TAP Lab.

Space Domain Awareness

Task, track, and account for activity across the space domain, turning fragmented sensor and provider data into a clear operating picture.

Satellite collection tasking and accountability
Physics-based orbital pass prediction with confidence scoring
Coverage-gap detection and root-cause analysis
Integration with production data layers

Signal Processing and Intelligence

Real-time processing of sensor and signal data to detect, characterize, and localize events across distributed sensor networks.

Distributed sensor network integration
Real-time event detection and triggering
Signature detection and identification
Low-latency streaming data pipelines

Distributed Systems Observability & Integration

Make multi-vendor systems observable and accountable, from data lineage to service grading. Proven at the USSF SDA TAP Lab.

Automated compliance and schema grading across services
End-to-end data lineage and message-chain tracing
Multi-vendor service integration over standard interfaces
Real-time operational visibility and validation

Artificial Intelligence & Machine Learning

Applied, explainable machine learning for prediction and decision support, with secure on-premise and air-gapped deployment.

Predictive analytics and forecasting
Explainable ML for GO/NO-GO decision support
Secure air-gapped, on-premise AI
Model development, validation, and deployment

Recent Publications

Sharing our expertise and findings with the broader scientific community.

Utilizing Civilian Launches as Live Exercises for Evaluating a Federated Protect and Defend SDA Battle Management System

Furlich, G., Crews, A., Rossland, S., Marino, J., et al. (including Prasad-Rao, J.)

The Advanced Maui Optical and Space Surveillance (AMOS) Technologies Conference, 2025

This paper describes how the USSF SDA TAP Lab used live civilian launches as end-to-end exercises to evaluate Welders Arc, a federated protect-and-defend prototype SDA battle management system built through three-month Project Apollo accelerator sprints engaging 100+ organizations across industry, academia, and government.

Space Domain Awareness Battle Management SDA TAP Lab Live Exercises AMOS

Machine learning-based surrogates for eVTOL performance prediction and design optimization

Rao, J. P., & Chimata, S. N.

Metascience Aerosp, Vol. 1, pp. 246-267, 2024

This research develops computationally efficient machine learning surrogate models trained on nearly 50,000 eVTOL simulations to rapidly predict vehicle endurance and range, replacing resource-intensive physics-based models for design optimization.

Machine Learning eVTOL Performance Prediction Design Optimization

Smart buildings in the IoT Era--necessity, challenges, and opportunities

Heidary, R., Rao, J. P., & Fischer, O. J. P.

Handbook of Smart Energy Systems, pp. 1-21, 2023, Springer

This chapter examines how IoT-enabled smart buildings can address urbanization and climate challenges, reviewing current sensing capabilities, applications, and the need for interoperability between smart buildings and smart cities.

Smart Buildings IoT Energy Systems Building Management

Detecting Manufacturing Defects in PCBs via Data-Centric Machine Learning on Solder Paste Inspection Features

Prasad-Rao, J., Heidary, R., & Williams, J.

arXiv preprint arXiv:2309.03113, 2023

This study demonstrates a data-centric machine learning approach to detect PCB manufacturing defects across 15,000 PCBs, combining pin, component, and PCB-level analysis to improve automated quality control in electronics manufacturing.

Manufacturing Defects PCB Inspection Machine Learning XGBoost Data-Centric AI

Attrition Risk and Aircraft Suitability Prediction in US Navy Pilot Training Using Machine Learning

Prasad-Rao, J., Fischer, O. J. P., Rowe, N. C., Williams, J. R., Puranik, T. G., Mavris, D. N., Natali, M. W., Tindall, M. J., & Atkinson, B. W.

Aerospace, Vol. 10, Issue 4, pp. 379, 2023, MDPI

This research uses machine learning on 10 years of U.S. Navy pilot training data to predict aviator attrition risk and recommend suitable aircraft types, achieving 50% attrition prediction accuracy with 4% false positives and potential savings of USD 20 million annually.

Machine Learning Pilot Training Attrition Prediction U.S. Navy Aircraft Suitability

A Reusable Framework for Fault Detection and Isolation in Small Satellites

Rao, J. P., Pace, J., Williams, J., Mackey, R., & He, L.

37th Annual Small Satellite Conference, 2023

This NASA-supported research presents a reusable, data-driven framework using machine learning and Dynamic Time Warping to detect and isolate satellite anomalies in real-time, achieving 90% detection rates with less than 1% false positives by monitoring correlations between battery metrics and system operational variables.

Small Satellites Fault Detection Machine Learning Anomaly Detection NASA

Unsupervised anomaly detection using batteries in electric aerial vehicle propulsion test-bed

Pace, J., Rao, J. P., Williams, J., & He, L.

Annual Conference of the PHM Society, Vol. 14, Issue 1, 2022

This research presents BDAV (Battery-based Diagnosis for Aerial Vehicles), an unsupervised anomaly detection framework that uses battery measurements as a root of trust to diagnose other vehicle subsystems through machine learning models that capture physical dependencies between battery and operational variables.

Anomaly Detection Electric Aircraft Battery Systems Unsupervised Learning Vehicle Diagnostics

Experimental study into optimal configuration and operation of two-four rotor coaxial systems for EVTOL vehicles

Rao, J. P., Holzsager, J. E., Maia, M. M., & Diez, J. F.

Aerospace, Vol. 9, Issue 8, pp. 452, 2022, MDPI

This experimental study analyzes coaxial rotor configurations for eVTOL vehicles, finding that optimal performance in two-rotor systems requires sequential operation (back motor first), while four-rotor systems achieve maximum efficiency with gradually increasing propeller pitch values from first to last rotor.

eVTOL Coaxial Rotors Experimental Study Propulsion Systems Performance Optimization

Capabilities & Research