SBIR-STTR Award

Platform Agnostic Data Storage Infrastructure
Award last edited on: 2/25/2022

Sponsored Program
SBIR
Awarding Agency
DOD : SOCOM
Total Award Amount
$900,406
Award Phase
2
Solicitation Topic Code
SOCOM201-001
Principal Investigator
Noah Jacobsen

Company Information

Mission Solutions Group Inc

6655 Jet Park Road Suite 102
North Charleston, SC 29406
   (843) 481-3131
   N/A
   www.missionsolutionsgroup.com
Location: Single
Congr. District: 06
County: Charleston

Phase I

Contract Number: N/A
Start Date: 8/27/2020    Completed: 8/6/2021
Phase I year
2020
Phase I Amount
$1
Direct to Phase II

Phase II

Contract Number: 6SVL4-20-C-0003
Start Date: 8/27/2020    Completed: 8/6/2021
Phase II year
2020
Phase II Amount
$900,405
In this Phase II work program, Mission Solutions Group will develop and deliver to USSOCOM a software prototype system for the Platform Agnostic Data Storage Infrastructure (PADSI) Small Business Innovation Research (SBIR) program. Our technical solution is designed to provide a common framework for multi-model, multi-layer data storage and import/export control as well as advanced Artificial Intelligence and Machine Learning (AI/ML) services for the Special Operations Forces Data Layer (SOF-DL), for a plurality of end-user applications ranging from Common Operational Picture (COP) in fully connected command and control environments to warfighter Situational Awareness (SA) in resource challenged tactical environments. Our technical solution, based on the Lambda web-services architecture, delivers near-real time performance with advanced AI/ML processing on streaming data inputs combined with long-term analytics derived from historical data stores. Our solution builds upon the basic Lambda architecture using robust open source component technologies to provide several key innovations: (1) a multi-model graph database to host and query multi-format, multi-layer data within the same database object, enabling efficient extraction of cross-layer data and interoperability with diverse data-formats and standards, (2) support for multiple leading ML tool-kits for executing pre-trained Neural Network (NN) models enabling vendors to easily deploy their proprietary models into the SOF-DL with a common AI/ML service architecture, (3) support for ML Development Operations (DevOps) enabling vendors to rapidly train and validate new models with a pre-approved, standardized development pipeline, and (4) an Application Programming Interface (API) that provides a standard set of commands for applications in the SOF-DL for data import/export control and AI/ML model execution. Our Phase II proposal includes a technical collaboration with MIT Lincoln Lab's AI Software Architectures and Algorithms Group, via a subcontracting agreement. Lincoln Lab, a Federally Funded Research and Development Center (FFRDC), brings deep talent and experience developing and delivering advanced AI/ML prototypes to DoD customers, to this SBIR program. Specifically, their Open Architecture for Artificial Intelligence (OA4AI) platform shares similar design objectives with the PADSI SBIR program, and our planned collaboration will result in the development and delivery to USSOCOM a high impact integrated OA4AI-PADSI Prototype system, as well as integration, demonstration, and delivery of at least one AI/ML application already developed by Lincoln Lab.