NASHVILLE (TSU News Service) – A major problem U.S. military counter-insurgency operations face is the ability to easily identify hostile enemy group intent and hidden dangers in obstructed environments.
Such group activities are generally embedded in clutters in urban locations, involving well-trained individuals who blend in with the general population to carry out their mission. The results usually are surprise attacks and high civilian casualties.
A Tennessee State University mechanical engineering professor thinks he has the answer. As a result of a proposal to the U.S. Army Research Office, he has won a $334,000 defense grant to investigate the possibility of developing an advanced technology that improves the capability of automated surveillance systems.
Dr. Amir Shirkhodaie, professor in the College of Engineering and director of the Center of Excellence for Battlefield Sensor Fusion, said his research will develop a new capability for behavioral pattern learning of partially obscure group activities that take place in confined, obstructed spaces.
“The ultimate goal of this project is to develop a robust information-theoretic framework with supportive techniques that can detect obscure group activities in areas such as inside a vehicle, boat, airplanes or corner alleys of urban areas,” said Dr. Shirkhodaie.
He said this could greatly reduce the false alarm rates in surveillance operations that frequently occur as a result of miscalculation of enemy intent, and help shift the “balance of power” in peacekeeping operations.
“If we can deliver this kind of technology to the battlefield, this is a game-changer,” said Maj. Jay Deason, an aviator with the Tennessee Army National Guard, who has served two tours in Iraq, flying Black Hawk Helicopters.
He said while this technology would have limited application for air reconnaissance operations, it would be greatly useful to ground forces and civil affairs specialists, who identify critical requirements needed by local citizens in combat or crisis situations.
Civilians would also greatly benefit from this technology in homeland security, crowd control, and anti-drug and anti-crime operations, Dr. Shirkhodaie said.
Maj. Deason, who has also served one tour along the southwest U.S. border flying UH-72 helicopters, said this technology will greatly help the civilian population and in border patrol operations.
“This is very exciting. This technology has the capability to save lives,” Deason added.
The main objective of Dr. Shirkhodaie’s proposal, “Detection of Partially Observable Group Activities (POGA) in Confined Obstructed Spaces,” is to develop context-based taxonomy and ontology schema for coherent analysis and inferences of POGA.
The investigation will take place in three phases, including the development of a robust Adaptive Image Processing technique for detecting and tracking of behavior pattern of POGA; a Computational Intelligence technique based on a hybrid neuro-fuzzy system architecture; and a Multi-Layer Hidden Markov Model technique for probabilistic spatiotemporal state transition modeling that leads to context-aware discovery on anomalous group activity.
In student learning, Dr. Shirkhodaie said the project would greatly enhance research opportunities for TSU students in this area, as well as offer scholarly training opportunities for underrepresented minority students in the STEM disciplines.
The Dean of the College of Engineering, Dr. S. Keith Hargrove who also announced three new research projects with Boeing for more than $500,000, congratulated Dr. Shirkhodaie on his award, adding that the grants represent the dedication and commitment of faculty to research and attracting students to the College of Engineering.
“The opportunity for academia and industry to collaborate to solve industry problems makes the company more competitive, and enhances the quality of our engineering programs for students and faculty,” said Dr. Hargrove.
The three Boeing projects include using artificial intelligence for the development of aircraft propulsion controls; the development of resilient control mechanisms to mitigate cyber attack in engineering embedded systems; and the development of mathematical models for energy harvesting and storage.
The faculty members involved in these projects are Drs. Sachin Shetty, Mohammed Saleh Zein-Sabatto, both professors of Eleectrical Engineering; and Dr. Landon Onyebueke, professor of Mechanical Engineering.