LITTLETON, Colo., 9 Sept. 2005. PercepTek Inc. was recently awarded two Air Force contracts to provide research and development on using unmanned aerial vehicles (UAV's) in critical autonomous avionic operations.
There are several hundred UAVs currently flying in Iraq, and UAVs are an important component of the Army's $15 billion Future Combat Systems project. The rest of the armed forces expect to spend an additional $15 billion on UAV technology over the next 10 years. These air platforms could greatly benefit from the two Air Force programs awarded below.
The first contract was issued by the U.S. Air Force Office of Scientific Research (AFOSR) in Arlington, Va. and is for "Joint 3D Reconstruction of Static Background and Moving Targets Using Structure from Motion."
The Air Force has a compelling need for better automatic target recognition algorithms, for operation in complex dynamic environments where both opposing force and non-combatants are present. As an aircraft flies over a possible target, it must distinguish between friendly, enemy, and neutral vehicles with as little human involvement as possible. Current approaches either require sensors which are too heavy, too big, or too easy to detect, or they rely on overly restrictive assumptions about the scene, such as assuming only stationary targets.
The overall goal for this project is to develop a system that can take video collected from a moving air platform, along with some amount of motion information, and recover the shape of the terrain, as well as the motion and 3D structure of any objects that are moving through the scene. PercepTek has chosen to team with the Colorado School of Mines, PLACE ???? for this project.
The second contract was issued by the Air Force Research Laboratory (AFRL) at Eglin Air Force Base in Eglin, Fla. and is for "IR Algorithms for Video-Aided Navigation."
An important aspect of autonomous UAV operations is a robust navigation capability. UAVs need to traverse extended (several kilometer) trajectories over time periods up to several hours. Robust autonomous navigation under these conditions requires successful operation with and without GPS. Many autonomous development systems today rely on the consistent availability of GPS to implement their navigation solution. However, GPS dropouts are not uncommon, typically occurring next to buildings or occluding land formations. GPS dropouts can range from short (on the order of seconds or minutes) to indefinite in duration when operating in a GPS-denied area.
Equally important is the requirement for robust navigation during both day and night conditions. This project will develop a FLIR-based visual odometry capability for UAVs, giving them this robust autonomous navigation capability. PercepTek is again teamed with the Colorado School of Mines for this project.
"While PercepTek has made its mark in the unmanned ground vehicle business over the past 5 years, we've gained a great deal of attention lately with the recent award of the Army's Aviation Applied Technology Director's UAV program, and now these 2 new Air Force projects," said Steven Paulet, PercepTek's director of business development.
PercepTek, Inc., of Littleton, Colo., a leader in the field of intelligent robotics, is focused on integration, perception, planning and control for the robotics, surveillance and autonomous vehicle markets. Aggressively pursuing research and development with the commitment to develop and transition advanced technology products, PercepTek works with government sponsored research programs, collaborates with key university and research institutes, and develops strategic long-term relationships with industrial partners to assist in broadening product lines and incorporating automation technologies. For more information, see www.perceptek-robotics.com.