DIA Opens Major AI and Machine Learning Funding Opportunity

Facebooktwittermail

DIA artificial intelligence machine learningThe Defense Intelligence Agency is getting ready to spend big on artificial intelligence and machine learning companies with the opening of a wide ranging funding opportunity.

DIA officials posted 17 separate AI and machine learning technology solicitations on their website, NeedipeDIA, to make it easier for startups and small businesses to access DIA acquisition programs. DIA connected the solicitation to a funding vehicle called a BAA that is designed to streamline the acquisition process.

Program leaders will host an Industry Day on Aug. 2-3 at DIA headquarters at Joint Base Anacostia-Bolling in Maryland, to offer feedback on the solicitations. The event is open to academics and industry.

The solicitations can be found at NeedipeDIA, which also breaks down how to apply for funding from the BAA and offers a great list of hints on how to do it. For example, initiatives costing under $650,000 have the best opportunity to be funded through the BAA.

DIA is the largest intelligence agency in the government as it includes Defense Department and intelligence community (CIA, NSA) agencies.

We copied the 17 solicitations on NeedipeDIA specific to AI and machine learning and posted them below:

99.0:  Using Machine Learning and Natural Language Processing (NLP) to Identify and Organize Weapon System Information in Big Data

Description:  DIA seeks to understand industry capabilities and methodologies for leveraging Machine Learning techniques combined with other tools like NLP to automatically identify technical terms and names associated with complex descriptions of weapon systems buried in various types of data sources.  These data sources may include raw sensor files, unstructured electronic documents, and various types of multimedia files.  These files also may come in various formats.  Certain file types may require other pre-processing tools like optical character recognition (OCR) in addition to NLP.  Military systems are inherently complex, and weapon systems are often described in various places using a series of letters and numbers and associated “nick names.”   Foreign language descriptions, when present, add another layer of complexity.  Tools and methods should be able to determine out how to manage dialect and media recognition models associated with weapon systems that are highly unique to the intelligence and military communities.  These tools and methods must be able to be applied to and then work with most open source NLP software applications, search tools, and other expert systems.  Once organizational expertise has been distilled into knowledge systems that support AI tools, these knowledge systems have to be able to be used to support other AI tools.  Once relevant data has been identified, this data will become part of the updated, centralized data model describing the weapon system.  Responses to this need can address all or part of this description.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.1:  Use of AI and Machine Learning Tools to Conduct Collection, Research, Information Monitoring, Automated Reporting, Thematic Data Management, Data Transformation, and Database Development

Description:  DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to conduct all source collection, research, information monitoring, automated reporting, and database development.  DIA seeks tools that streamline and accelerate certain intelligence workflows; to compress intelligence production timelines; to integrate multiple data sets and formats; to identify relationships, connections, correlations, and associations between data; to process data and discern relevance with respect to a given topic; to facilitate information sharing; and to prepare data for anticipatory analysis.  These tools must act on large amounts of unstructured data in various formats and potentially accommodate dynamic data flows.  AI and machine learning would also support the streamlining and acceleration the more time consuming/intensive tasks of analyzing unstructured information, audio, images/visualizations and video.  DIA also seeks to understand industry capabilities for applying artificial intelligence and machine learning to thematic data management and data transformation in a way that allows a semi-automated approach to the extract-transform-load (ETL) burden for collected data.  Specific ETL functions include the ability to select, focus, simplify, tag, and transform overtly or covertly collected data into human or machine interpretable form for further analysis or action.  Responses to this need can address all or part of this description.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.2:  Machine Learning Support to Workflow Automation Activities

Description:  DIA seeks to understand industry capabilities and methodologies for using machine learning tools to automate intelligence production, planning, and process workflows to reduce the time spent accomplishing these tasks manually.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.3:  Tools for Predictive Analysis, Alerting, and Indications and Warning (I&W)

Description: DIA seeks to understand industry capabilities and methodologies for applying artificial intelligence, machine learning, and predictive analysis algorithms to Indications and Warning (I&W) within a big data framework.  This need will identify semi-automated tools that can conduct analytics of open source data, intelligence sensor data, finished intelligence, financial intelligence, and other forms of intelligence reporting to locate trends, create alerts that require attention, and to recommend actions based upon sensor feeds.  Solutions will identify modeling techniques, and will analyze and visually display predictive analytics on changes observed in the battlespace and areas of control during peacetime and during conflict.  Specific focus areas include technologies that 1) can manage the expert input often required to make I&W tools leveraging AI and Machine Learning; 2) will have the capability to “learn” from complicated and diverse information flows that are used for I&W; and then 3) will interface with users in a useful way to achieve I&W outcomes (visualization of data, tasking, exploitation, analysis, alerts, and decision support).  Responses to this need can address all or part of this description.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.4:  Semi-Autonomous Multi-Sensor Fusion Leveraging Artificial Intelligence (AI)

Description: DIA seeks to understand industry capabilities and methodologies regarding technology that can leverage AI-enabled sensor processors to manage and fuse multiple sensors of the same or different phenomenology and then to recognize and respond to signature sources by type and to also identify anomalies.  The AI enabled sensor network will then predict and respond to recognized threat types and threat activity based upon these sensory feeds.  Sensory feeds may incorporate analog and digital feeds from acoustic, seismic, magnetic, density/pressure, electromagnetic, radio frequency, electro-optic/infrared, hyperspectral, and other domain signatures correlated in both space and time.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.5:  Artificial Intelligence and Machine Learning Support to Military Operations

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of fusing Multi-INT sensor data for real-time battlespace awareness and predicative analysis at the strategic, operational, and tactical levels of warfare.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.6:  Artificial Intelligence (AI) and Machine Learning Support to Business Operations

Description: DIA seeks to understand industry capabilities for applying AI and machine learning to the area of business operations to include acquisition management; financial analysis; portfolio prioritization and optimization; business analytics; risk management; resource conservation; and business decision support.  The desired solution would enhance the ability to streamline and gain insight through predictive analysis of business financial operations.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.7:  Artificial Intelligence and Machine Learning Support to the Data Science Environment

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of a data science environment.  CIO is looking for recommendations to identify capabilities that support and streamline Natural Language Processing (NLP) of text and audio data, Recommender Engines, Net Flow Data, and features of the Data Science Environment.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.8:  Artificial Intelligence and Machine Learning Support to Finished Intelligence Products and Knowledge Management

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to scan all finished intelligence products and create from them a body of organized knowledge that people can use to explore topics and concepts. The body of knowledge should return an answer to the user that describes the connections between the data up to the time the user entered the request for information.  The solution should return de-duped and de-conflicted data sources to provide analysts a comprehensive end product with accurate sourcing.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.9:  Artificial Intelligence and Machine Learning Support to Open Source Information Gathering

Description: DIA seeks to understand industry capabilities for using AI and machine learning tools to semi-autonomously collect all forms of open-source information and then updated analytical techniques for data mining and discovery.  Solutions should include the ability to combine, compare, and analyze classified and open source material and attempt to cross-verify information obtained from different sources.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.10:  Machine Learning Support to the Management of Tasking from Multiple Sources

Description: DIA seeks to understand industry capabilities for applying machine learning technologies to the target area of processing incoming tasking from multiple sources, particularly email.  This capability prevents tasking from entering an organization without a formal tasking process and helps ensure that ad hoc requirements are tracked and captured by managers.  The solution would leverage machine learning tools read email message flows to discover, track, and route tasks to official tasking channels.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.11:  Artificial Intelligence and Machine Learning Support to Dynamic Threat Analysis Modeling

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of dynamic threat analysis modeling.  Solutions would identify an automated system that will allow analysts to develop weighted analytical threat models leveraging quantified adversary capabilities (either using analyst input or by pulling from existing databases), doctrine, influence, historical behavior, and stated intent.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.12:  Artificial Intelligence and Machine Learning Support to the Assessment of Performance Measures

Description DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of assessing organizational performance measures.  Solutions will develop a universal performance measurement capability across the Defense Intelligence Enterprise (DIE) to support CCMD and agency priorities.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.13:  Artificial Intelligence and Machine Learning Support to Human Resource Recruiting

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of human resource recruiting.  Solutions will be a capability that correlates structured and unstructured data sources to identify, recommend and rank individuals to fill unique billets by leveraging personal biographies, expertise profiles, and available online data.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.14:  Artificial Intelligence and Machine Learning Support to Publication Author Relationships

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of evaluating structured and unstructured data contained within publications to show relationships between authors on a specific topic.  Capability should leverage natural language processing to auto tag and build networks based on relationships between authors, coauthors, institutions, topics, and other details.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.15:  Artificial Intelligence and Machine Learning Support to Signature Identification

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of identifying signatures from various sensor sources to aid in exploitation, analysis and production.  Signature identification supports Ballistic Missile Technical Collection, Nuclear Monitoring, and MASINT capabilities.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.16:  Artificial Intelligence and Machine Learning Support to Records Management

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of front end and back end management of records, business analytics, and visualization.  Identify a talent management (personnel) system leveraging machine learning to manage, track, analyze, trend, and forecast employee career enhancement opportunities.  The system should contain a web-based user input for clients to enter and verify their data.  The system should contain relational tables to Human Resources (HR) data and employee training records.  The system tables should allow talent managers to manage/track and forecast training courses, assignments, leadership tracks, promotion actions, Joint Duty Assignments (JDA’s), stretch assignments or moves that impact employee’s development and other opportunities.  The system should also contain a section/table to house employee career counseling records conducted by talent managers.  Additionally, the system should contain a business analytics/visualization dashboard to quickly portray the health and welfare of the employee workforce.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

99.17:  Artificial Intelligence and Machine Learning Support to Cybersecurity

Description: DIA seeks to understand industry capabilities for applying artificial intelligence and machine learning to the target area of cybersecurity. Solution would have the ability to dynamically detect anomalies/risks/threats and then to alert defenders.  The capability must be able to perform these functions at a greater scale and pace than what is currently accomplished using the traditional collect, process, and analyze methodology.

• Response Type: BAA
• Status: Open/Considering – all responses due by 17 July for potential presentation at DIA’s Industry Outreach Days 2-3 August 2017
• Additional Evaluation Criteria: None

Leave a reply

Basic HTML is allowed. Your email address will not be published.