Session Sheet - 2026 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2026): Defense Manufacturing I

2026 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2026): Defense Manufacturing I
Program Organizers: David Leigh, University of Texas at Austin

Tuesday 1:30 PM
August 4, 2026
Room: Zlotnick Ballroom 2
Location: AT&T Center

1:30 PM  Invited
Efficient AM Qualification: Lessons from the U.S. Navy and America Makes: Yash Parikh¹; ¹EOS of North America, Inc.
     Qualifying additively manufactured components for defense applications remains one of the most significant barriers to scaling additive manufacturing (AM) across the Defense Organic Industrial Base (OIB), particularly for submarine production, sustainment, and fleet readiness. This presentation connects two urgent priorities: accelerating laser powder bed fusion (LPBF) qualification and expanding a trained, production-ready AM supplier base for the U.S. Navy’s maritime and submarine industrial base. Using EOS’ work with the U.S. Navy Maritime Industrial Base initiative as a central example, the presentation examines how AM is advancing from isolated demonstrations toward deployable industrial capability. It highlights the role of supplier development, workforce training, production discipline, and repeatable qualification methods in building a more resilient maritime manufacturing ecosystem.The technical core of the presentation introduces a data-driven qualification framework that integrates in situ process-monitoring signals, including Optical Tomography, Melt Pool Monitoring, and Powder Bed imaging, to create a digital record of LPBF builds. By linking real-time process signatures to final material performance, AI-enabled models can support faster qualification, reduce reliance on destructive testing, and help manage the transition from single-laser to multi-laser production platforms. This approach aligns with broader defense qualification efforts, including America Makes’ Delta Qualification work, where EOS led a project focused on rapid qualification pathways for LPBF in critical applications. Together, these examples show how coordinated industry, government, and standards-based efforts can shorten qualification timelines, strengthen the AM supplier base, and accelerate the deployment of qualified parts for Navy and defense applications.

2:00 PM Question and Answer Period

2:10 PM
Rapid Health Monitoring Artifacts for Process Qualification and Failure Risk Assessment in Laser Powder Bed Fusion: Albert To¹; ¹University of Pittsburgh
    Process anomalies and powder contamination in laser powder bed fusion (L-PBF) remain key barriers to qualification and reliable deployment in defense applications. This work introduces compact, geometry-informed rapid health monitoring artifacts that provide fast, in situ indicators of process stability, material performance, and failure risk. The first artifact quantifies ductility through residual-stress-driven cracking in an inverted L-shaped cantilever with a notched interface. Crack length correlates strongly with elongation in notched tensile tests for Inconel 718, enabling rapid assessment of material performance and sensitivity to powder moisture. The second artifact enables in situ detection of process instability through a mid-build fracture event governed by residual stress and material strength. The fracture layer provides a simple, scalable metric sensitive to spatial variation and laser degradation, supporting accelerated qualification and improved reliability in defense manufacturing.

2:30 PM
Panel Discussion: Navigating the Complexities of Advanced Manufacturing for National Security: Lonnie Love¹; Mohan Karulkar¹; Nadine Miner¹; ¹Sandia National Labs
    Advanced Manufacturing is reshaping the production of critical components, yet national security applications require exceptional levels of reliability and assurance. This panel brings together experts from diverse backgrounds to share their perspectives on the unique challenges facing advanced manufacturing in defense contexts. Discussion topics will include supply chain transparency, digital data integrity, qualification and certification challenges, and the interplay between cyber and physical systems. Panelists will explore issues such as variability in manufacturing processes, detection of latent defects, and the alignment of emerging technologies with stringent defense requirements. Attendees will gain insight into the broad technical, operational, and policy considerations that influence the adoption of advanced manufacturing for high-consequence national security systems. This session aims to foster a rich dialogue on the complexities and evolving landscape of trustworthy manufacturing for defense applications.

2:50 PM
Real-Time Object Detection for Autonomous Robotic 3D Printing: AGYA ABOAGYE-OTCHERE¹; Jianzhi Li¹; Ahmed Bendaouia¹; ¹Institute for Advanced Manufacturing
    Robotic 3d printing is a growing section within the additive manufacturing field. Robotics and AI in manufacturing have provided the platform to complete autonomy within these systems. This research aims to develop a robust computer vision system for real-time object detection to 3d print with a 6DOF robot in varying light conditions. The models are trained on custom datasets to perform defect detection and part identification. This vision-based data is mapped to the robot’s motion planning algorithm for autonomous adjustments. The implementation of the object detection system allows for precise monitoring of the printing process, enabling the robot to identify components and detect structural anomalies like warping and layer shifting. The model trained on all light conditions produced the best performance with 0.895 mAP50 and 0.880 mAP50-95. The best lighting-restricted model reached 0.830 mAP50 overall and slightly exceeded the baseline on the environmental subset alone with 0.936 mAP50 versus 0.924.

3:10 PM
RocketSmith: An Agentic System for High-Powered Rocket Design and Manufacturing: Peter Pak¹; Jesse Barkley¹; Rumi Loghmani¹; Derek Baich²; Ananya Pamal¹; Amir Barati Farimani¹; ¹Carnegie Mellon University; ²Tripoli Rocketry Association
    This work presents RocketSmith, an agentic system capable of the design, manufacturing, and iteration for the development of high powered rockets. The system enables the intelligent automation of software tools as to not only validate factors such as flight stability but also generate the parametric design components for the rocket assembly. A collection of subagents and skills enable the optimization of flight parameters via iteration in both zero-shot and human-in-the-loop workflows. With this system, four distinct high power rockets with various motor and assembly configurations were developed utilizing the unique design capabilities of additive manufacturing. These assembly components were fabricated using various FDM printers, manually evaluated for flight readiness, and flight tested at a launch event. From these tests, the two of the four rockets were successfully recovered in reflyable condition and collected flight data established an 85% altitude accuracy to performed flight simulations.

3:30 PM
Heating Rate–Driven Interfacial Stability in Cold-Sprayed SS316L on AISI 4140: Abishek Kafle¹; Santosh Rauniyar¹; Kalen Baker¹; Ben Xu¹; Wei Xiong²; Weihang Zhu¹; ¹University of Houston; ²University of Pittsburgh
    CSAM repair of steel components appears highly promising, but heat treatment of different metals can be rather difficult owing to coefficient of thermal expansion (CTE) mismatch. In this study, SS316L was deposited on AISI 4140 steel, and influence of heating rate during post-deposition heat treatment (PDHT) on interface behavior was investigated. Results demonstrate heating rate dependency: samples heated at a rate of 300°C/h showed interfacial cracking and delamination, whereas those heated at 80°C/h survived such conditions. Microstructural analysis revealed metallurgical bonding and mechanical interlocking in the intermetallic layers. To investigate possible reasons for heating rate dependence, a simplified finite element model was employed, and stress evolution during heating was analyzed. As a result, it can be assumed that interface response depends on interaction between thermal mismatch stress and stress relaxation. Thus, it is evident that heating rates during heat treatment are crucially important for cold-sprayed different metals interface survival.

3:50 PM
Advanced Manufacturing for National Security: Nadine Miner¹; Mohan Karulkar¹; Lonnie Love¹; ¹Sandia National Labs
     While Advanced Manufacturing (AdvM) has matured to enable novel tools, materials, and geometries, the national security enterprise faces a persistent barrier: how to trust what AdvM produces. In low-volume, high-consequence production, confidence diverges from industry norms. Metrics like scale, lifetime, and data transparency do not align across sectors, and qualification schemas fail to accommodate AdvM’s hallmarks such as stochastic defect formation and incomplete process–structure–property understanding.This talk presents a risk-informed technical perspective on where trust breaks down across the national security product realization lifecycle and what it will take to rebuild it. Drawing on curriculum developed jointly with UT Austin and Sandia’s CAMINO and Academic Alliance initiatives, it offers a practical framework to close a critical disconnect the field can no longer defer.

4:10 PM
Benchmarking U-Net, FCN-ResNet50, and DeepLabV3-ResNet50 for Outdoor AMR Scene Segmentation: MD AL AMIN MEIA¹; Md Shahriar Forhad¹; Monsuru Ramoni¹; Jianzhi Li¹; ¹University of Texas Rio Grande Valley
    We present a comparative evaluation of U-Net, FCN-ResNet50, and DeepLabV3-ResNet50, for outdoor autonomous mobile robot navigation. The experiments were conducted using an outdoor mobile robot segmentation dataset containing RGB images and pixel-level segmentation masks with four semantic classes: Road, Sidewalk, Cycle Path and Other. The images and masks were resized to 192×512 pixels, and all models were trained for 200 epochs in high-performance computer. U-Net was implemented as a custom encoder-decoder segmentation architecture, while FCN-ResNet50 and DeepLabV3-ResNet50 used residual backbones for dense pixel-wise prediction. The models were evaluated using precision, recall, F1-score, Dice score, IoU, class accuracy, mean IoU, IoU threshold metrics, and computational cost. Based on the class-wise full-test evaluation, DeepLabV3-ResNet50 achieved the highest mIoU of 0.9207, followed by FCN-ResNet50 with 0.9040 and U-Net with 0.8314. These findings indicate that residual backbone-based segmentation models provide stronger and more consistent performance for outdoor AMR scene understanding.