Acoustic Training Data Prioritization
ID: N241-030Type: BOTH
Overview

Topic

Acoustic Training Data Prioritization

Agency

Department of DefenseN/A

Program

Type: SBIRPhase: BOTHYear: 2024

Topic Link

https://www.sbir.gov/node/2479739

Additional Information

https://www.defensesbirsttr.mil/SBIR-STTR/Opportunities/
Timeline
    Description

    The Department of Defense (DOD) is seeking proposals for the topic of "Acoustic Training Data Prioritization" as part of the SBIR 24.1 BAA. The Navy branch is specifically interested in developing a tool that assesses training data for artificial intelligence or machine learning (AI/ML) algorithms used in detecting and tracking submarines. The current paradigm of using large sets of data for training AI/ML algorithms is costly and may not yield optimal results. The Navy is looking for a tool that can analyze acoustic data collected by undersea warfare systems and prioritize data that is diverse, representative, and as small as possible for training AI/ML algorithms. The tool should reduce the amount of training data used while maintaining or improving performance, as measured by the Receiver Operating Characteristic (ROC) curve. The Phase I of the project involves developing a concept and demonstrating feasibility using unclassified data, while Phase II focuses on designing and delivering a prototype tool for testing and evaluation. The technology developed in this project has potential applications in various industries that rely on AI/ML training.

    Files
    No associated files provided.
    Similar Opportunities
    DOD SBIR 24.4 Annual - Novel Positioning, Navigation, and Timing (PNT) Signal Classification Techniques
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for a novel positioning, navigation, and timing (PNT) signal classification techniques. The purpose of this solicitation is to develop the capability to classify signals in real-time that impact navigation systems. The goal is to better understand the type(s) of signal(s) experienced in relevant environments to apply mitigation techniques before harm can be done. The research will focus on improving performance, cost savings, and expanding the application of the technology sensor solution set to include additional Army aviation assets. The project will involve developing adaptive learning techniques using AI/Machine Learning (ML) algorithms to perform PNT signal classification. The proposed solution aims to build upon the progress made in AI/ML signal classification and move towards real-time signal classification. The project will have a Phase I and Phase II, with Phase II requiring the development of two antenna systems capable of detecting and classifying interference signals in real-time. The project also aims to make the antenna design portable to support upgrading antenna systems and providing support to other antenna systems in the same environment. The anticipated duration of the project is until March 31, 2025. For more information and to submit proposals, visit the DOD SBIR website.
    DOD SBIR 24.4 Annual - xTechScalable AI 2
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the xTechScalable AI 2 topic. This solicitation focuses on two main areas: Scalable Tools for Automated AI Risk Management and Algorithmic Analysis: The Army is looking for automated tools to evaluate and mitigate risk against an AI Risk Management Framework (RMF) to ensure trusted and validated AI models. The tools should be able to evaluate multiple dimensions of AI risk, classify and quantify AI risk, and propose mitigation options. The Army is particularly interested in tools that can accept risk-related inputs from multiple data sources and modalities and have standardized evaluation methods and mitigation strategies. Scalable Techniques for Robust Testing and Evaluation (T&E) of AI Operations Pipelines: The Army needs a robust and automated T&E approach for AI Operations Pipelines. This includes evaluating data integrity, data labeling methodologies, and model training. The Army is interested in tools that can identify and evaluate data integrity, assess the quality and accuracy of labels applied to training datasets, and evaluate model performance in terms of resource consumption, robustness, scalability, and privacy and security. The Phase I proposals can receive up to $250,000 for a 6-month period, while Direct to Phase II proposals can receive up to $2,000,000 for an 18-month period. Phase II involves producing prototype solutions that will be evaluated by soldiers, and Phase III focuses on maturing the technology and producing prototypes for further development and commercialization. The xTechScalable AI 2 prize competition will be used to identify small businesses eligible to submit proposals under this topic. The full solicitation can be found at this link.
    DOD SBIR 24.4 Annual - xTechScalable AI
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the topic of "xTechScalable AI" as part of the SBIR program. The Army branch is specifically interested in novel and disruptive concepts and technology solutions that can address the vulnerabilities of current machine learning pipelines and models. The goal is to develop comprehensive security models that can defend against universal AI threat vectors and are scalable to rapidly evolving threats. The Army is particularly interested in proposals that focus on systematic testing and evaluation methods, trusted and secure validation and verification strategies, continuous monitoring capabilities, improved transparency and assurance of code and data, and improved telemetry capabilities. The Army will prioritize submissions from winners of the xTechScalable AI prize competition. The project will involve a Direct to Phase II submission, followed by the production of prototype solutions that are easy to operate by soldiers. Phase III will focus on maturing the technology to TRL 6/7 and producing prototypes for further development and commercialization. The deadline for proposal submission is March 31, 2025. For more information and to submit a proposal, visit the solicitation agency's website at [solicitationagencyurl].
    DOD SBIR 24.4 Annual - xTechScalable AI
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the topic "xTechScalable AI" as part of the SBIR program. The Army branch is specifically interested in novel and disruptive concepts and technology solutions that can address the vulnerabilities of current machine learning pipelines and models. The goal is to develop comprehensive security models capable of defending against universal AI threat vectors. The Army is prioritizing proposals that focus on systematic testing and evaluation methods, trusted and secure validation and verification strategies, continuous monitoring capabilities, improved transparency and assurance of code and data, and improved telemetry capabilities. The Army will use the xTechScalable AI prize competition to identify small businesses that meet the criteria for award, and only winners of the competition will be eligible to submit a proposal under this topic. The project will have three phases: Phase I involves submitting a Direct to Phase II (DP2) proposal, Phase II involves producing prototype solutions for evaluation by soldiers, and Phase III involves completing the maturation of the technology and producing prototypes for further development and commercialization. The deadline for proposal submission is March 31, 2025. For more information and to submit a proposal, visit the solicitation agency's website at [solicitationagencyurl].
    DOD SBIR 24.4 Annual - Novel Positioning, Navigation, and Timing (PNT) Signal Classification Techniques
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for a novel positioning, navigation, and timing (PNT) signal classification techniques. The purpose of this solicitation is to develop the capability to classify signals in real-time that impact navigation systems. The goal is to better understand the type of signals experienced in relevant environments to apply appropriate mitigation techniques. The current navigation systems depend on radio frequency (RF) signals that can be influenced by various interference sources. The challenge is to quickly understand the signal characteristics to react and mitigate negative impacts. The proposed solution aims to build upon AI/Machine Learning (ML) algorithm technologies to perform PNT signal classification in real-time. The project will involve developing two antenna systems capable of detecting and classifying interference signals, collecting relevant signals for training the AI/ML solution, and demonstrating the ability to detect and identify signal types in a relevant environment. The project will have a Phase I and Phase II, with Phase II focusing on the development and demonstration of the antenna systems. The anticipated duration of the project is until March 31, 2025. For more information and to submit proposals, visit the DOD SBIR website.
    DOD SBIR 24.4 Annual - Chief Digital and Artificial Intelligence Office (CDAO) Data Mesh Reference Design (REFDES)
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the Chief Digital and Artificial Intelligence Office (CDAO) Data Mesh Reference Design (REFDES). The objective is to make all data across the Department of Defense seamlessly interoperable while retaining federated control, hosting, and ownership. The DOD aims to establish a set of software services that allow data users to discover, access, and consume DoD data products via self-service API. The data access must support attribute-based access control (ABAC) and operate in a zero-trust environment. The program consists of three phases: Phase I involves creating a formal REFDES, Phase II focuses on building a Minimum Viable Product (MVP) version of the design, and Phase III involves delivering a full production capability. The REFDES must address key concepts such as VAULTIS compliance, services communication model and framework, data templating, dynamic attribution association, automated notification services, cybersecurity, and zero trust support. The resulting data mesh will support interoperability for applications in both battlefield and boardroom settings. The solicitation is open until March 31, 2025. For more information, visit the SBIR topic link or the solicitation agency website.
    DOD SBIR 24.4 Annual - xTechScalable AI 2
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the xTechScalable AI 2 topic. This solicitation focuses on two main areas: Scalable Tools for Automated AI Risk Management and Algorithmic Analysis: The Army is looking for automated tools to evaluate and mitigate risk against an AI Risk Management Framework (RMF). The tools should be able to evaluate multiple dimensions of AI risk, classify and quantify AI risk, and propose mitigation options. The Army is particularly interested in tools that can accept risk-related inputs from multiple data sources and modalities and have standardized evaluation methods and mitigation strategies. Scalable Techniques for Robust Testing and Evaluation (T&E) of AI Operations Pipelines: The Army needs a robust and automated T&E approach for AI Operations Pipelines. This includes evaluating data integrity, data labeling, and model training. The Army is interested in tools that can identify and evaluate data integrity, assess the quality and accuracy of data labels, and evaluate model performance in terms of resource consumption, robustness, scalability, and privacy and security. The Phase I proposals can receive up to $250,000 for a 6-month period, while Direct to Phase II proposals can receive up to $2,000,000 for an 18-month period. Phase II involves producing prototype solutions that will be evaluated by soldiers, and Phase III focuses on maturing the technology and producing prototypes for further development and commercialization. The xTechScalable AI 2 prize competition will be used to identify small businesses eligible to submit proposals under this topic. The full solicitation can be found at the following link: [solicitationagencyurl].
    DOD SBIR 24.4 Annual - Chief Digital and Artificial Intelligence Office (CDAO) Data Mesh Reference Design (REFDES)
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the Chief Digital and Artificial Intelligence Office (CDAO) Data Mesh Reference Design (REFDES). The objective is to establish a set of software services that enable seamless interoperability of data across the Department of Defense (DoD) enterprise while retaining federated control, hosting, and ownership. The DoD aims to break the data out of stovepipes and create a data mesh that allows data users to discover, access, and consume DoD data products via self-service API. The data access must support attribute-based access control (ABAC) and operate in a zero-trust environment. The program consists of three phases, starting with the development of a formal REFDES in Phase I, followed by the creation of a Minimum Viable Product (MVP) in Phase II, and concluding with the delivery of a full production capability in Phase III. The REFDES must address key concepts such as VAULTIS compliance, services communication model and framework, data templating, dynamic attribution association, automated notification services, and cybersecurity and zero trust support. The resulting data mesh will support interoperability for applications in both battlefield and boardroom settings. The proposal submission deadline is March 31, 2025. For more information, visit the solicitation link.
    DOD SBIR 24.4 Annual - User and Entity Behavior Analysis
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the topic of "User and Entity Behavior Analysis" as part of their SBIR program. The objective of this research is to develop a UEBA capability that serves as a Policy Decision Point (PDP) in the Tactical Zero Trust Architecture (ZTA). The UEBA will analyze user and entity behavior by collecting activity data and applying advanced analytics to detect anomalies. The solution will leverage data already collected and normalized by the Elastic Stack and incorporate various sources such as Active Directory, endpoint systems, firewalls, and vulnerability scans. The UEBA should include a flexible REST API for obtaining telemetry and making authorization decisions. The project will be conducted in three phases. Phase I requires a proof of concept in the form of a whitepaper, demonstrating the feasibility of developing the UEBA capability. Phase II involves developing a prototype to collect and interpret data, display risk score changes, and allow human decision-making based on alerts. Phase III focuses on dual-use applications, such as embedding AI/ML pattern recognition into cybersecurity operations and applying UEBA to IoT, healthcare, and finance sectors. The project duration is not specified, but the solicitation is open until March 31, 2025. For more information and to submit proposals, interested parties can visit the DOD SBIR website at [solicitationagencyurl].
    DOD SBIR 24.4 Annual - Direct to Phase II: Next-generation Autonomy for Unmanned Maritime Vehicles (UMVs)
    Active
    Department of Defense
    The Department of Defense (DOD) is seeking proposals for the topic of "Direct to Phase II: Next-generation Autonomy for Unmanned Maritime Vehicles (UMVs)" as part of the SBIR program. The Navy is specifically interested in developing next-generation autonomy that will increase the capability and scope of utility of UMVs while decreasing the level of remote human operator involvement. The goal is to enable UMVs to perform complex tasks with little to no human intervention, handle dynamic and harsh maritime environments, support diverse missions and tasks, and cooperate autonomously with other UMVs. The technology should operate with low-bandwidth and intermittent communication and be robust to uncertain and inaccurate perception information. The project will involve the development of algorithms and software, simulation-based testing, and in-water testing with physical-small scale models. The Phase II period of performance is anticipated to be four years. Successful completion of the project could lead to the transition of the Next-generation Autonomy software to the acquisition program for use within its architecture.