TECHNOLOGY AREAS: Air Platform, Sensors, Battlespace

OBJECTIVE: Develop and demonstrate innovative methods for determining and predicting the four-dimensional trajectory of an aircraft given limited sensor data.

DESCRIPTION: Naval Safety Center accident data indicate that mid-air collisions remain in the top-five causal factors for Class-A accidents in Naval Aviation. In addition, numerous organizations are struggling to introduce Unmanned Aerial Systems (UAS) into the National Air Space (NAS) for reconnaissance, patrol, and other security and law enforcement missions. The integration of UAS into the NAS is being met with resistance from the Federal Aviation Administration (FAA) due, in part, to the lack of a "sense and avoid" capability in the UAS � a means of being aware of the airspace around the UAS and the ability to avoid conflict with other air vehicles.

A solution revolves around knowing where the "threat" is and being able to predict the trajectory of that "threat" in realtime. The threat may be anything from a data-link equipped military aircraft capable of transmitting a near-complete picture of its aircraft state (position, velocity, acceleration, etc.) to a general aviation aircraft incapable of providing any state information beyond its position. Position information may be available from on-board radar, FAA ground-based surveillance radar via Automatic Dependent Surveillance-Broadcast (ADS-B), or on-board Mode-S transponders, for example. Development of advanced trajectory prediction algorithms utilizing limited sensor data will be critical to increasing UAS mission capability as general aviation aircraft state information is normally only available from aircraft or ground-based radars.

In order for proposed solutions to be effective, the errors associated with such a prediction must be understood and the error budgets incorporated into the prediction allowing for intelligent determination of the probability of imminent mid-air collision. For example, large potential errors increase the range at which a collision avoidance warning must be issued to be effective and not nuisance-prone. Such errors may include effects from aircraft sensors and radar, ground-based radar, communication latencies, aircraft maneuvering capabilities, and pilot intention based upon recent trajectory history.

PHASE I: Develop the concept for a realtime predictive methodology and demonstrate the scientific merit and feasibility of the approach when given near-complete aircraft state data and when given position only. Identify the conditions under which the effectiveness of proposed predictive methods would be affected (i.e., range, closure rate, level of maneuvering, etc.).

PHASE II: Fully develop the methodology demonstrated under Phase I into a usable predictive tool. Evaluate the accuracy of predicted trajectories against provided threat data. Identify the conditions (i.e., range, closure rate, etc.) under which the errors cause the prediction to fail leading to erroneous collision warnings. Determine the accuracy of the trajectory predictions against the actual trajectories. Modify prediction algorithms to allow for multiple data sets to include steady-state flight, general aviation-like maneuvering flight and high performance flight.

PHASE III: Refine and deliver algorithms for threat prediction for numerous data sets and sources. Transition the technology to various defense platforms.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: Improved capability in mid-air collision predictions leading to fewer nuisance warnings, increased user acceptance, and integration of unmanned aerial systems into the National Airspace.

REFERENCES:
1. Naval Safety Center Class "A" Mishap Data, 1997-2004.

2. "New algorithms for aircraft intent inference and trajectory prediction", YEPES Javier Lovera, HWANG Inseok, ROTEA Mario, Purdue University, AIAA Journal of Guidance, Control, and Dynamics, 2008

3. "Performance Evaluation of a Novel 4D Trajectory Prediction Model for Civil Aircraft", Marco Porretta, Marie-Dominique Dupuy, Wolfgang Schustera, Arnab Majumdara and Washington Ochieng, Journal of Navigation, Cambridge University Press, 2008.

KEYWORDS: Mid-Air Collision Avoidance Systems; Sense and Avoid; Detect and Avoid; Unmanned Aerial Systems; Trajectories; Sensors