SBIR-STTR Award

LG-based Generation of COAs, Short- and Long-term plans, as well as Interactive 5-D Overlays for the MTC2/C2X Planning System
Award last edited on: 6/4/2021

Sponsored Program
SBIR
Awarding Agency
DOD : Navy
Total Award Amount
$1,224,998
Award Phase
2
Solicitation Topic Code
N181-015
Principal Investigator
Vladimir Yakhnis

Company Information

Stilman Advanced Strategies LLC (AKA: STILMAN)

501 South Cherry Street Suite 1100
Denver, CO 80246
   (303) 717-2110
   info@stilman-strategies.com
   www.stilman-strategies.com
Location: Single
Congr. District: 01
County: Denver

Phase I

Contract Number: N68335-18-C-0550
Start Date: 6/1/2018    Completed: 11/29/2018
Phase I year
2018
Phase I Amount
$125,000
In complex tactical scenarios, mission planners are often confronted with mission goals and operational environment restrictions that limit their options. High cognitive workloads produced by manual assessments and operational concerns increase tactical decision timeframes. To address these problems and support mission planning and operations within National and Foreign Airspaces, Space, Maritime, and Land domains, mission planners need automated planning assistance which can quickly animate (visualize) the routes and mission plans and the tactical risks to personnel and property. To maintain the tactical initiative, Mission planners and Battle Management Aids must optimize mission performance in faster than real-time under dynamic, multi-variable conditions allowing Mission Commanders to initiate tactical actions on their terms. LG-RAID addresses these shortcomings. Using LG-Artificial Intelligence (AI) as a multi-domain mission planner, LG-RAID constructs an optimal 4D, geo-rectified, animated overlay that provides entity visualization and predictive Course of Action analysis capability for dynamic, multi-variable conditions within seconds on a standard desktop computer.

Benefit:
(1) What is the first non-DoD product that this technology will go into? This product will be useful in the commercial / civilian delivery, and Law Enforcement (LE) domains as a UAS Planner and possible controller. (2) Who will be the customers, and what is the estimated market size? This market is estimated to exceed 20K licenses for LE delivery applications should exceed 100K licenses in 5 years. (3) How much money will be needed to bring the technology to market, and how will that money be raised? Approximately $6 million and 2 years, would be needed to fully model the LE and delivery domain including their unmanned systems. These funds would be raised by using commercial credit (loans), research investments from LE, and product sales. (4) Does the company contain marketing expertise and, if not, how will that expertise be brought into the company? Yes, STILMAN has in-house marketing expertise. Additional support can be provided by consultants and sub-contractors on a case-by-case basis. (5) Who are the proposing firms competitors, and what is the price and/or quality advantage over those competitors? There are no direct competitors providing an LG-based solution. If they entered the market tomorrow, their ramp up and learning curve would be extremely steep. It is anticipated that it would take at least 4 or 5 years (and $25 $30M) to produce an equivalent capability.

Keywords:
Intelligent Agents, Intelligent Agents, Linguistic Geometry, Modeling and Simulation, Artificial Intelligence, 4D visualization, Behavior modeling optimzation, mission rehearsal

Phase II

Contract Number: N68335-21-C-0152
Start Date: 12/10/2020    Completed: 12/16/2022
Phase II year
2021
Phase II Amount
$1,099,998
To provide enhanced situational awareness and understanding in multi-domain environments, new methods to generate COAs as well as short- and long-term plans are needed. Artificial Intelligence (AI) can enable the above COA/plans and Weapon-Target-Pairing generation which, in turn, will provide automation for enhanced situational awareness and understanding in multi-domain environment. Spreadsheets and 2-D representations have proven to be effective in identifying unit support, assigning roles, tasks and actions within the maritime, air and land mission domains, however they are limited in their ability to visually represent multi-unit or multi-domain temporal coordination. Whereas the traditional AI techniques, when applied to military domains, are either non-scalable or are uncapable of explaining (using notions that the users could understand) the generated solutions (e.g., plans or decisions), we propose to a novel application of the new Explainable AI called Linguistic Geometry (LG). Given a complex mission in multi-domain environment, LG-AI rapidly generates COAs as well as short- and long-term plans with their 2-D or 3-D (whichever appropriate, as selected by users) visualizations. Combined with capability to animate those COAs and plans (i.e., providing additional dimension representing battlespace dynamics) this provides an enhanced situational awareness/understanding in multi-domain environment. LG AI technology is scalable and, as it generates plans and COAs, it simultaneously generates their explanations in terms of Actions-Reactions-Counteractions that are intuitively graspable by military personnel. These dynamic, predictive COAs have been verifiably shown to improve the temporal coordination of assets and quantify decision and reaction times as well as increasing the situational understanding of tactical performance envelopes and tactical tradeoffs. 4-D and 5-D representations emerge when 2-D and 3-D spatial representations are combined with two additional dimensions: (1) temporal (enabled by animation representing battlespace dynamics) and (2) parameter spaces such as dynamic threats, distributed sensor and weapon coverage areas, and other key mission factors, such as decision and reaction times. This will be provided via LG-AI based 5-D overlays.

Benefit:
Benefits to the US Navy: For AEGIS Common Baseline, Maritime Tactical Command and Control (MTC2), LG-RAID will provide an ECOA-based mission planning service and real-time Course of Action (COA) generation capability for MTC2 as a networked service . This capability will become an integrated part of NAVSEA PEO IWS 1.0, Decision Aid Tools, see Transition Roadmap figure below. In addition to providing a multi-domain distributed mission planner and real-time ECOA planning capability to enhance manned and un-manned systems control, LG-RAID will also provide an automated or supervisory control capability for real-time weapon-target pairing recommendations for long-range targets and in-close immediate reaction time engagements. Through the established MTC2 on-boarding process, LG-RAID will be vetted through fleet users during major exercises and the AEGIS Combat System Testing and AEGIS Training Centers (schoolhouses). This two-stage performance review process will confirm user acceptance and accelerate transition. The key elements (highlights) of this process were previously described in Section 1.2. As described, the process will provide continuous insight into LG-RAIDs performance through IPRs, PMRs and software testing at NAVSEA, Dahlgren, SPAWAR and within major Fleet exercises. Benefits to the commercial markets This system can be used by Commercial Security Firms, Law Enforcement, and DHS in municipal and Federal Law Enforcement applications. The system can ingest large amounts of data and perform automatic planning, coordination and correlations of sensors in monitoring specific targets. This capability will decrease military I law enforcement/ first responder planning and analysis times in using sensing assets as well as the time required to complete the mission. Additionally, this system will enable law enforcement to increase the competency of the threat (e.g., scale the threat competency up or down) for training and

Keywords:
Linguistic Geometry, multi-domain missions, LG hypergames, Surface Action Group, multi-threat COA analysis, automated predictive planning, enhanced decision support, Adversarial reasoning