ARES Security and Bevilacqua Research Sign Agreement to Launch AI Decision Support Software

Nov 25, 2019 | News & Articles

ARES Security and Bevilacqua Research Corporation (BRC) have signed a Strategic License and Alliance Agreement to further develop artificial intelligence software solutions that meet the market demands of DoD, Government, Commercial and Critical Infrastructure customers who require rapid decision support to respond to major incidents and complex threats. Under the agreement, ARES Security will develop commercial software for rapid decision support, called AVERT AI, based upon BRC’s Cognitive Object Reasoning Engine (CORE) machine learning technology that has been proven in U.S. Army deployments over the last 20 years. ARES Security has acquired a global license for this technology, exclusive in the Physical Security Market and plans to productize these technologies and integrate them into the AVERT Platform.

Threats are evolving at an accelerated pace making the ability to quickly identify and appropriately respond increasingly challenging. Experience and in-depth training are essential to quickly make the right decision for an effective response. But it is nearly impossible to train and prepare for each potential threat scenario and even more unlikely to have the experience required to respond to emerging sophisticated threats. Together, ARES Security and BRC will market AVERT AI and build solutions that improve the rapid decision-making process that is increasingly critical to incident response when seconds count. ARES Security will also cross-license the entire AVERT suite of solutions to BRC that include Assessment, Training, C2, and AI products.

Ben Eazzetta, CEO of ARES Security noted: “The security challenges that we face today require a new way of thinking. ARES Security is committed to providing innovative and disruptive security solutions that allow our customers to find new ways to increase the effectiveness of their security plans, optimize their security and operational spending, and improve incident response. ARES Security is pleased to have the opportunity to work with Andy Bevilacqua and his BRC team to bring AVERT AI to market and enable rapid decision support in critical situations.”

ARES Security is the developer of the AVERT suite of physical security solutions. In 1999, ARES Security started work with DTRA on the physical security vulnerability assessment for the US stockpile of nuclear weapons using modeling and simulation software. Over the years, the AVERT software has matured in many ways. AVERT Physical Security has been Accredited by DoD and DOE having gone through the rigors of Verification and Validation. AVERT Physical Security has also been Certified under the Safety Act by the Department of Homeland Security. Today, AVERT Physical Security continues to be used by DTRA to protect the US Nuclear Weapons Stockpile, the Pentagon Force Protection Agency, and 65% of the commercial nuclear reactors in North America. The capabilities of AVERT are also used by Transit Agencies, Critical Infrastructure, and Corporations to optimize security designs and improve response plans for challenging threats such as active shooter, IED, and vehicle as a weapon. What has been proven through many years of DoD, DOE, and Commercial work is that AVERT Physical Security can significantly increase the effectiveness of customers’ security plans while saving millions of dollars in capital and operational costs. ARES Security extends the capabilities of AVERT Physical Security to provide virtual reality exercise and training solutions for Security Forces to acquire the skills needed for incident response. As well as AVERT C2 that fuses the data from the dozens of systems that are prevalent in Emergency Operations Centers today into a Common Operational Pictures to provide the situational awareness that is so critical for rapid security incident response.

Andy Bevilacqua, CEO / CTO of BRC, commented, “Although BRC has been able to use the CORE tools to provide specialized Artificial Intelligence and Machine Learning solutions for many different DoD programs over the last 2 decades, this is the first time we have licensed this unique capability for use by another company. This collaboration with ARES Security not only provides them with a mature, leading-edge technology that gives them clear leadership in the security product market, it also provides BRC with valuable feedback from a community of users that will allow us to continually improve our technology”.

BRC is a VA-certified Service-Disabled Veteran-Owned Small Business, headquartered in Huntsville, AL, with satellite offices in Niceville and Panama City, FL and Edwards AFB, CA. BRC has been in operation for over 27 years and is an ISO 9001:2015 certified company with approximately 350 employees located in 17 states. Since 1996, BRC has been at the forefront of improving human behavioral models using a set of tools that would allow developers to create machine learning algorithms without the need to be experts in either artificial intelligence or computer learning. The Cognitive Object Reasoning Engine (CORE) Machine Learning Toolkit was first developed by BRC in 1996 under an Army Small Business Innovative Research (SBIR) program. Today, the technology includes:


  • The Conceptual Graph Builder/Editor. This tool allows a subject matter expert to sit down at the computer and input his/her knowledge into the computer using an easy-to-use graphical editor based upon conceptual graphs (Sowa,1984).

  • An Adaptive Resonance Theory (ART) Neural Network Tool. This tool allows real machine learning to be implemented quickly by the team. ART neural networks (Carpenter & Grossberg,1987), unlike classic neural networks, do not destroy previously learned information when new information is added. This allows ART Neural Networks to operate in much the same way that the human brain works during the learning process (see Miller,1956, Baddeley and Hitch,1974).

  • Various visualization Tools for the Visualization of ART Data Clusters Used in Machine Learning. These tools allow multi-dimensional data to be visualized when identifying ART data clusters during the initial training process.