Guided Systems is focused primarily on the application of its patented neural network adaptive control methods in the aerospace sector. Its technology is used to dramatically reduce the time and money required to complete guidance and control system design, development and validation, and also offers the potential to significantly increase a systems tolerance of faults or failures when redundant means for actuation are available. While applicable to manned flight systems, missiles, munitions and spacecraft, the technology is particularly well suited to the development programs typical of unmanned flight vehicles, and is especially effective in application to rotorcraft and other types of complex vertical take-off and landing vehicle designs. The technology also offers great benefits for control of flexible structures, including the flexible airframes that are characteristic of next generation High Altitude Long Endurance (HALE) aircraft designs. Adaptation is accomplished by augmenting a tradition linear or nonlinear controller design with an artificial neural network. The network requires no preflight training. The benefits of neural network adaptation include both the ability to tolerate a large degree of model uncertainty and unmodeled dynamics. Dependence on wind-tunnel data and high-fidelity dynamic modeling in support of control system design is greatly reduced or even eliminated, as is the need for gain scheduling. Other benefits include control system design tolerance of configuration evolution, fault and damage tolerance, and the ability to adapt in flight to dramatic configuration changes (such as due to weapons launch, pick-up/drop-off of a large payload, or battle damage). GST originally applied the cited neural network adaptive control technology in several high profile DoD applications in partnership with Boeing. These included the 1998 Air Force/Boeing/GST Reconfigurable Systems for Tailless Fighter Aircraft (RESTORE) program flight demonstration on the X-36 (with on-going transition to the DARPA/Boeing UCAV), and an Air Force/GST/Boeing/Georgia Tech 2001 Advanced Adaptive Autopilot flight test on the JDAM (Joint Direct Attack Munition). The technology developed in this latter program has entered production on the MK-82 JDAM in 2007. Guided Systems has focused much of its attention on the unique control problems of rotorcraft. The companys experience with automatic control of unmanned rotorcraft spans more than a decade, and includes early success in the Association for Unmanned Vehicle Systems Internationals Aerial Robotics Competition (winning Georgia Tech team in 1993), flight demonstration of early neural network research (neural network adaptive control for the U.S. Armys Free Flight Rotorcraft Research Vehicle 1994-95), funded programs for demonstration of technologies for autonomous flight operations in remote monitoring applications (Department of Energy Demonstration at the Savannah River Nuclear Reactor Facility, 1995) and surveillance (U.S. Army Autonomous Scout Rotorcraft Testbed program, 1996), and demonstration of very high-bandwidth neural network adaptive guidance and control technology on a Joint GST/Georgia Tech unmanned rotary wing testbed (U.S. Army support through Georgia Tech Rotorcraft Center from 1997-2001). More recently the companys adaptive control architecture has been applied to a tilt rotor, a ducted fan, on a coaxial helicopter, and on the Skytote rotary wing UAV. (See the section on Past Programs for a more detailed summary of these latter programs). In 2007, Guided Systems entered a formal partnership with Cloud Cap Technology as their rotorcraft solution provider, making this unique rotorcraft expertise and control system technology available in finished product form.
