Our problem sets come from the Department of Defense and Intelligence Community. 

Joint Improvised-Threat Defeat Organization (JIDO)

PROBLEM: Augmented Reality for Facial Recognition - How might we use Augmented Reality for U.S. security elements in order to identify threat individuals in crowds in unstructured environments using facial recognition.


U.S. personnel have difficulty identifying and tracking individuals in unstructured crowd environments.  Although centralized surveillance systems can scan crowds, mobile individuals and teams lack wearable technical means for autonomous surveillance. Personnel using Augmented Reality (AR) technology could scan crowds for individuals on a known list (of limited size) while the individual wearing a technical device is focused on another function or task. US forces could benefit from a technology that enables them to constantly scan crowds in structured or unstructured environments to identify individuals that may pose a threat.


US personnel cannot identify persons of interest in crowds.  Determine how Augmented Reality and facial recognition might assist individual operators.


·      Size of the database the system is referencing (200 person database nominal minimum, ideally database in tens of thousands)

·      Area being scanned

·      Size of the area that can be scanned through the glasses at any given time

·      Speed of the facial recognition software to access an individual’s profile (under 5 sec. nominal minimum)

·      Any additional equipment that needs to be used/accessed/worn by the individual

·      Stability/durability of the glasses and any other necessary equipment


Joint Improvised-Threat Defeat Organization (JIDO), Lieutenant Brennon Ducote, J8/TSD Special Programs brennon.p.ducote.mil@mail.mil/(703) 602-8924.


Asymmetric Warfare Group (AWG)

PROBLEM: Understanding Social Media & information Warfare - The U.S. Army currently lacks the ability to aggregate, index, and analyze large amounts of social media data in order to identify current and future political, social and economic trends. This valuable source of data could be used in conjunction with advanced artificial intelligence methods to inference future social unrest and allow commanders to focus resources more effectively on probable hotspots in dense urban environments.


Social media use has exploded worldwide. As the costs of mobile communication technology plummets, social media has become the preferred means of information exchange worldwide. Volumes of data now exist about historical trends and precedents in social media, providing vital, untapped clues to future social activity.


U.S. Army commanders currently lack the ability to aggregate, index, and analyze large amounts of social media data in order to identify probable hotspots of social unrest in dense urban environments.


·      Allow commanders to focus resources effectively

·      Aggregate social media “big data”

·      Extract sample data that is representative of larger population

·      Index social trends and statements indicative of trends

·      Index social media “Influence Factors,” “Influencers” and “Followers”

·      Analyze large data streams through machine learning and Boolean Logic

·      Harness Artificial Intelligence to identify social trends

·      Create automated system to identify and graphically depict potential future trouble zones in near real time

·      Disseminate generated predictive analysis information at the tactical level


U.S. Army’s Asymmetric Warfare Group (AWG), Kyle E. Hardy (kyle.e.hardy.ctr@mail.mil)


PROBLEM: Detecting Aerial-Based Threats


ISR and kinetic attacks via unmanned aerial systems is no longer achievable by military forces. Now any non-state actor can buy a drone on Amazon.com for well under $1,000 capable of military grade ISR missions and direct attack.


How might end users detect and identify small unmanned aerial systems, including commercially available fixed wing and rotary wing drones, in order to gain situational awareness of imminent, aerial based threats?


·      The solution should be man-portable, passive and easily trainable.

·      Commercially available drones should be the focus, not foreign military capabilities

·      Any environment worldwide (i.e. do not let rules and regulations of the FCC or available US-based frequencies limit solutions)

·      Open to any solution: electronic jamming, GPS jamming, GPS spoofing, software vulnerabilities, kinetic attack, EMPs etc.

·      Prefer the project be done at the academic level and NOT classified


US Army’s Asymmetric Warfare Group, Mark N. Holden, mark.n.holden.ctr@mail.mil.


U.S. Special Operations Command (SOCOM)

PROBLEM: Tactical One-Way Cross Domain Solution (CDS) - Rapidly enable unclassified tactical ISR sensor data ingestion directly into SIPRNet through connected tactical SECRET computer systems or other approved devices.


Current approved method of tactical data transfer requires end-users to air-gap sensor data to optical media, virus scan the sensor data on both network systems, before manually uploading to Secret Internet Protocol Router Network (SIPRNet) connected computer systems.  This method is not operationally viable in tactical situations and environments by SOF personnel.  Air-gapping is typically not being performed due to the complexity and time involved.  Potentially valuable and time sensitive tactical Intelligence, Surveillance, and Reconnaissance (ISR) sensor data is not being saved or archived and is unavailable to other SOF entities removed from the local tactical site.


Rapidly enable unclassified tactical ISR sensor data ingestion directly into SIPRNet through connected tactical SECRET computer systems or other approved devices.


Describe technical thresholds: Need a low cost, small form factor, USB connected, one-way (Unclass to SIPRnet) CDS.  CDS device needs to be capable of reading numerous removable media format types, removable media format sizes, other USB devices and the ability to connect analog and digital streaming video feeds.
Environmental Conditions: CDS should be able to operate in the same environment of any tactical computer system.
Technologies that might be relevant: Commercially available USB card readers, one-way diodes, anti-virus software, user authentication, device registration and auditing capabilities. 
Other clarifying information: Data ingestion will automatically be virus scanned. If nothing is found, the data will be released for further tactical processing and/or available for dissemination to other locations.  The sensor data comes from closed unclassified government sources. Tactical sensor data will include streaming video, still images, thermal images, radar images, acoustic files, magnetic signatures/strengths, office automation files or any other sensor data. CDS must have all the necessary security authorizations and approvals to connect to SIPRNet.


U.S. Special Operations Command Acquisition, Technology, & Logistics. 

Thomas (Brendon) Reedy, Thomas.Reedy@socom.mil, (813-826-3060)