US Air Force Multi Sensors Data Integration for Autonomous Sense and Avoid

MS Unmanned Systems - Embry Riddle Aeronautical University

7.4 - Research Assignment: Sense and Avoid Sensor Selection 

  US Air Force Multi Sensors Data Integration for Autonomous Sense and Avoid

US Air Force, through its sponsored Northrop Grumman Corporation, have developed and tested autonomous systems with sense and avoid (SAA) capacity. The design is scalable and based in “a comprehensive sensor suite comprising Traffic Alert and Collision Avoidance System (TCAS) and Automatic Dependent Surveillance - Broadcast (ADS-B) for detecting cooperative intruders as well as radar and electro-optical (EO) sensors for detecting non-cooperative intruders” (US Air Force, 2011).

The research is focused on the integration of data of the sensors, as part of the whole sense and avoids system.   

Integration of sensor data

“Multi-Sensor Integrated Conflict Avoidance (MuSICA) and consists of four major modules: Sensor Input Management (SIM), Sensor Data Integration (SDI), Jointly Optimal Conflict Avoidance (JOCA), and Flight Control Interface (FCI)” (USAF, 2011)

 

The module SIM consists in a pre-processing stage from the sensor of SA; the module SDI receives the data from the SIM and after processing, the output is an integrated tracking data about intruder. The module JOCA receives the processed data from SDI to analyze the situation and generate maneuver commands if it’s necessary to control the collision avoidance. These commands are received by the FCI which acts as a human pilot over the Flight control system of the vehicle.  The architecture of the SAA shows that the sense is addressed by the modules SIM and SDI, and the modules JOCA and the flight control interface addresses the avoid process.

Characteristics of the sensors for SAA

The suite of sensors for SAA involves different types of sensors such as “cooperative and non-cooperative as well as active and passive” (USAF, 2011). Cooperative sensors are considered omnidirectional and includes sensors TCAS and ADS-B, Non-cooperative are considered directional and at the same time can be divided in active (Radar/ LIDAR) or passive (EO/IR). “TCAS is an airborne secondary surveillance radar (SSR) system with surveillance and collision avoidance functions; An ADS-B-equipped aircraft broadcasts its own position and associated accuracy and integrity information to other ADS-B-equipped aircraft and ground receivers” (USAF, 2011).

Normally, the positions of UAVs are determined by a Global Positioning System (GPS) and altimeter by sensors of pressure; therefore, ADS-B is not strictly required for all national space of the USA. TCAS expects for acknowledgements from the receiver vehicle unlike the ADS-B, but both are cooperative sensors. At commercial level, the active radar sensors for UAVs are not available yet, but some laboratories such as MIT Lincoln Laboratory, has developed the “Airborne Sense and Avoid (ABSAA) Radar Panel” (MIT, 2014). EO/IR sensors are non-cooperative passive sensors which are mainly used for ground operations. 

 SDI- Objective of the Design

The objective of the design of sensors data integration, as its name states, is the integration of the multiple sensors with dissimilar characteristics with the purpose to provide a high precision and hardy tracking of intruder for conflict and avoidance of collision. Sub goals or specific objectives are focused on integration of the all-suite of sensors, generation of integrated tracking, getting better “features of dissimilar sensors in terms of accuracy, integrity, continuity, and availability (AICA) in the integrated tracks, and Achieve modular and efficient real-time software implementation” (USAF, 2011).

SDI - Architecture

Sense data integration architecture is composed of four functional modules “Extended Kalman Filter (EKF), Data Association, False Track Filtering (FTF), and Track Manager“(USAF, 2011).

The internal flow starts with the input of measurements from multiple sensors to the Track Manager; the module EKF is the responsible to propagate the tracking which were predicted and update them by addition of the updated tracking; the data association module synchronize measures with predicted tracks; then, the Track Manager analyze and establish the new data base of tracks and the initiation of new tracks or termination of old tracks; the FTF track establishes if the detection or sensing of some track is true or false; and finally the output of the updated track is declared to the function of avoidance of the collision.

 References

MIT Lincoln Lab's Airborne Sense and Avoid (ABSAA) Radar Panel Wins R&D Award - UAS VISION. (2014). Retrieved July 18, 2016, from http://www.uasvision.com/2014/08/27/mit-lincoln-labs-airborne-sense-and-avoid-absaa-radar-panel-wins-rd-award/

Multi-Sensor Data Integration for Autonomous Sense and Avoid. (2011). Retrieved July 18, 2016, from http://arc.aiaa.org/doi/pdfplus/10.2514/6.2011-1479