DON26BZ01-NV030 TITLE: Artificial Intelligence and Machine Learning (AI/ML) for Additive Manufacturing (AM)
OUSW (R&E) CRITICAL TECHNOLOGY AREA(S): Applied Artificial Intelligence (AAI)
COMPONENT TECHNOLOGY PRIORITY AREA(S): Advanced Computing and Software;Advanced Materials;Sustainment
PROJECTED CMMC LEVEL REQUIREMENT: Level 2 (Self)
OBJECTIVE: Automate additive manufacturing (AM) through advanced computational techniques (i.e., artificial intelligence and machine learning [AI/ML], digital twins, etc.) to select optimal materials and manufacturing parameters to meet mission requirements in terms of component performance.
DESCRIPTION: AM has enabled new designs and rapid fabrication. However, there are no automatic tools available to computationally link across build platform to part performance. This SBIR topic seeks to leverage AI/ML, digital twins, and process simulation to select optimal materials and manufacturing parameters to meet rapidly changing mission requirements. A user should be able to input material type, part geometry, and AM system details into the prototype tools to automatically generate optimized build parameters along with accurate mechanical performance predictions.
While some tools in the current market can address part of this need, none are known which can integrate across the entire material lifecycle from pre-build to performance in a single ready-to-use package. The focus of this effort will be investigating legacy parts (i.e., obsolete castings and forgings) which need rapid production to avoid long lead times. Leveraging physics-informed AI/ML technologies and digital twins to optimize printing based on geometry and material properties will mitigate build defects and reduce post-processing while enabling performance prediction.
From a technical standpoint, the prototype tool(s) developed under this topic should seamlessly integrate across the component lifecycle, from initial design (or reverse engineering) to build parameter optimization to mechanical performance prediction in structural metals, to enable the user to accurately fabricate mission-critical components. The tool(s) must be part and AM build system agnostic to ensure scalability to multiple locations across the Navy’s manufacturing enterprise with various materials, systems, and performance requirements.
PHASE I: Define and develop a concept which leverages AI/ML, digital twins, and process simulation to select optimal materials and manufacturing parameters to meet rapidly changing mission requirements. Perform modeling and simulation with pointed physical testing for validation on a single component to demonstrate feasibility of the proposed concept. Required Phase I deliverables (in addition to the Contract Deliverables listed in the DON BAA instruction) will include a report on how the proposed concept will be expanded should the proposer be awarded a Phase II contract.
PHASE II: Expand the concept into full prototype tool development and validation using at least two additional components of different material classes and AM build systems. Demonstrate reduction in material fabrication time through automatic parameter generation while also reducing defect rates and material waste. Required Phase II deliverables will include:
a) A report on how the proposed concept can be expanded to other materials and systems not demonstrated in the Phase I and II taskings
b) Production of prototype tool(s) ready for delivery and demonstration at two U.S. Navy affiliated facilities.
PHASE III DUAL USE APPLICATIONS: Delivery of the final prototype tool(s) to U.S. Navy facilities will demonstrate the feasibility of the proposed solutions. Follow-on demonstrations to non-Navy participants will enable other DOW, DoE, government, and industry partners to ability to view the solution and continue transition to other facilities. The expectation is that the tool(s) will be leveraged by any organization in need of efficient digital tools to predict component performance based on manufacturing details.
REFERENCES:
KEYWORDS: Additive Manufacturing; AM; Artificial Intelligence; AI; Machine Learning; ML; AI/ML; Digital Twin
| 5/18/26 | Q. | The Phase I description requires "modeling and simulation with pointed physical testing for validation on a single component to demonstrate feasibility of the proposed concept." Could the Government clarify the intended scope of "pointed physical testing" for the Phase I single-component validation:
(1) Is it acceptable for the offeror to validate the Phase I concept against existing published or third-party experimental datasets for the single component - for example, the NIST Additive Manufacturing Benchmark (AM Bench) datasets - in lieu of the offeror physically printing and mechanically testing that component within the Phase I period of performance? (2) If the Government expects the offeror to physically print and test the Phase I component, is a single representative coupon or specimen geometry sufficient, or is a functional legacy-equivalent component (i.e., a casting/forging-replacement part) required for the Phase I validation? This clarification affects the Phase I cost structure and test-plan scope. Thank you. |
| A. | 1. If the proposer were to use existing data for Phase 1 validation, it would be expected that they demonstrate exactly how their AI/ML tool(s) apply to using that dataset.
2. For any physical testing, we would expect a part to be selected which would normally be cast/forged and instead is being made using AM technologies. The focus is on mechanical performance of the final component - not on the mechanical performance of a test coupon. |
|
| 5/18/26 | Q. | |
| A. | ||
| 5/16/26 | Q. | The topic description specifies pointed physical testing for validation on a single component during Phase I. Our proposed approach plans for this validation to be performed by the offeror at a commercial AM facility under subcontract. Could the TPOCs please confirm this approach is consistent with the topic's intent, or indicate whether validation at a Navy-affiliated facility is preferred or available within the Phase I period? |
| A. | Yes, a commercial AM vendor would be an appropriate location for physical validation builds. | |
| 5/15/26 | Q. | In the phase II decription, you have, " Expand the concept into full prototype tool development and validation using at least two additional components of different material classes and AM build systems". Does using an additional 2 different manufacturer of L-PBF machines satisfy this (AM build systems) requirement in phase II? |
| A. | Yes, that would technically address the different AM build systems request. But we are specifically looking for validation that the AI/ML tool(s) can work in different processes, so inclusion of non-LPBF (such as wire-arc DED, laser DED, powder-DED, etc.) in Phase 2 would be preferred should LPBF be the Phase 1 AM process. | |
| 5/13/26 | Q. | 1. Material class priority: Our current Phase I plan centers on 316L SS, given the depth of available process-structure-property data and its relevance to naval applications. Are there other structural metal systems - particularly high-strength austenitic steels or other alloy classes - that would be higher priority for the legacy parts problem you have in mind?
2. Legacy component type: The solicitation references obsolete castings and forgings with long lead times. Are there specific component categories (e.g., structural brackets, valve bodies, housings) that represent the most acute need at Carderock or across the broader Navy manufacturing enterprise? 3. Antenna components: One of our team members has had prior discussions with Navy stakeholders regarding AM-fabricated antenna structures. While we recognize this topic is scoped to structural metals and mechanical performance, we wanted to ask whether RF or antenna hardware is within the aperture of this solicitation or whether that would be better suited to a different topic. |
| A. | 1. Material class priority: No, we do not have specific material classes in mind for this call. All additive manufacturing (AM) materials that could replace castings/forgings are of interest.
2. Legacy component type: No, we do not have specific component classes in mind for this call. 3. Antenna components: That is likely more suited for a different topic as we are focused on castings/forgings (i.e., structures) replacement with AM. |
|
| 5/8/26 | Q. | We meet the requirements for CMMC lv1 and we have ITAR compliance, but the cost to maintain CMMC lv2 security alone outweighs the award for the SBIR. Does this SBIR require CUI making lv2 required? |
| A. | The requirement for this topic is CMMC Level 2 Self Assessment which has no cost requirement, other than the cost of time. This requirement is to be met by time of Phase I award. | |
| 04/28/2026 | Q. | If printer software does not slice, then are we going to work with STL files or some other format? (e.g. IGES) 8. What kind of compute capability will be available for AI or ML? |
| A. | We are expecting the AI/ML tool to be able to ingest all widely available commercial CAD file formats. As for compute capability, that will depend on the specific location it is deployed, but you should not assume that every location will have access to high-performance computing (HPC) capabilities. | |
| 04/28/2026 | Q. | If the printer software slices the project, then can we get access to sample output of the slices for training? |
| A. | That would be something for you to determine. This will not be provided by the Navy. | |
| 04/28/2026 | Q. | Will we be able to assume slicing and print-plans are going to be handled by the printer's software? |
| A. | Yes, that is a safe assumption for the LBPF systems, but not necessarily for all of the larger AM fabrication systems (such as those used in wire-arc DED). | |
| 04/27/2026 | Q. | When printing, is the ML algorithm expected to predict physical strength and durability characteristics? |
| A. | The AI/ML tool is expected to predict material performance (including strength, fatigue, etc.) based on the optimized material/process combination from the tool. | |
| 04/27/2026 | Q. | When printing, is the ML algorithm expected to predict physical strength and durability characteristics? |
| A. | The AI/ML tool is expected to predict material performance (including strength, fatigue, etc.) based on the optimized material/process combination from the tool. | |
| 04/27/2026 | Q. | How many different material types are expected to be managed by the AI? |
| A. | We would expect the AI/ML tool to be able to select from all metallic alloys currently used for AM fabrication (titanium, nickel, steel, stainless steel, copper, etc.) | |
| 04/27/2026 | Q. | How many AM printers is the AI supposed to manage? |
| A. | We would expect the AI/ML tool to be able to select from all metallic-based AM printers/systems currently used in the commercial AM marketplace. | |
| 04/27/2026 | Q. | Is the AI/ML component of this project focused on process optimization and path finding for the AM control and success? |
| A. | The AI/ML tool is expected to predict material performance (including strength, fatigue, etc.), which is directly related to the material, AM process, and selected process parameters used to create the part. | |
| 04/27/2026 | Q. | This topic requires validation using a single component in Phase I. Could you please advise whether you have any recommendations for the specific component-such as preferred materials, geometries, or other relevant details? Thank you. |
| A. | There are no recommended materials or specific components which will be provided. We do suggest validating using a component that would normally have been fabricated using casting or forging techniques. | |
| 4/15/26 | Q. | Does the Government have a preference for a specific modeling approach (e.g., physics-informed neural networks), or if offerors are encouraged to propose any general AI/ML methodologies that can meet the end-to-end integration performance objectives outlined in the topic? Additionally, to what extent is flexibility in model architecture and user-driven customization (e.g., tailoring models to different materials, geometries, and AM systems) viewed as a priority versus predefined or standardized modeling frameworks? |
| A. | There is not a preference for a specific modeling approach; the need is any AI/ML methodology that meets the performance objectives.
Secondly, model flexibility is paramount to enable performance prediction across various AM materials and processes. |
** TOPIC NOTICE ** |
The Navy Topic above is an "unofficial" copy from the Navy Topics in the DoW FY-26 Release 1 SBIR BAA. Please see the official DoW Topic website at www.dodsbirsttr.mil/submissions/solicitation-documents/active-solicitations for any updates. The DoW issued its Navy FY-26 Release 1 SBIR Topics pre-release on April 13, 2026 which opens to receive proposals on May 6, 2026, and closes June 3, 2026 (12:00pm ET). Direct Contact with Topic Authors: During the pre-release period (April 13, through May 5, 2026) proposing firms have an opportunity to directly contact the Technical Point of Contact (TPOC) to ask technical questions about the specific BAA topic. The TPOC contact information is listed in each topic description. Once DoW begins accepting proposals on May 6, 2026 no further direct contact between proposers and topic authors is allowed unless the Topic Author is responding to a question submitted during the Pre-release period. DoD On-line Q&A System: After the pre-release period, until May 20, 2026, at 12:00 PM ET, proposers may submit written questions through the DoW On-line Topic Q&A at https://www.dodsbirsttr.mil/submissions/login/ by logging in and following instructions. In the Topic Q&A system, the questioner and respondent remain anonymous but all questions and answers are posted for general viewing. DoW Topics Search Tool: Visit the DoW Topic Search Tool at www.dodsbirsttr.mil/topics-app/ to find topics by keyword across all DoW Components participating in this BAA.
|