Sunday, May 8, 2016

Obstacle Sense and Avoidance for Small Unmanned Aircraft



   In this blog we will explore the latest in sense and avoid technology that is commercially available for unmanned aircraft under 55 pounds. The ability for small unmanned aircraft systems (UAS) to perceive their surroundings and react in a in a timely and appropriate manner is critical characteristic not only for the safety of the aircraft itself, but also for manned aircraft sharing the airspace (Warwick, 2014). Current and emerging technologies are understandably placing a high degree of focus in implementing and refining UAS systems that can meet both requirements. In this paper, a commercially available sense and avoid system that aids in obstacle clearance will be examined.

   The ‘Guidance’ sensing module, offered by the DJI Corporation (DJI, 2016a), is arguably the most capable system available at this time for commercially available small UAS (Corrigan, 2015). The Guidance module consists of a suite of ultrasonic sensors and cameras which feed data to an onboard processor. Visual data is analyzed by means of an advanced computer vision algorithm that compares imagery from the cameras to determine an objects distance by comparing the relative perspective of the object from each camera (DJI, 2016b).

   The onboard processor consists of a pair of ARM Cortex-A9 cores which are able to provide up to 4000 Dhrystone million instructions per second while consuming less than 250 mW per CPU (ARM, 2016) making this device ideal for small UAS with limited power resources. The processor also incorporates various Field Programmable Gate Array (FPGA) hardware accelerators which consists of an array of programmable logic blocks (Xilinx Inc., 2016) to assist in rapid data processing. Such computational power supports data fusion from all installed sensors and automatically selects the best vision algorithm based on the camera configuration (DJI, 2016c).

   An article posted by Digital Trends (Digital Trends, 2016)  further describes the Guidance system as consisting of five ultrasonic sensors facing front, rear, left, right, and down that allows a Guidance-equipped UAS to determine its position relative to its surrounding with an accuracy of a few centimeters (Prindle, 2015). In addition, the Guidance also features a downward-facing visual positioning system, which aids in GPS-denied environments, that is effective up to 65 feet above the ground (DJI, 2016d). 

   A  Guidance core unit measures 78.5 mm x 53.5 mm x 14 mm with sensor measurements of 170 mm x 20 mm x 16.2 mm. The weight of a fully-equipped module (core plus five sensors with cabling) equals approximately 337 grams. Maximum power draw is 12 watts with all five sensors installed. Effective sensor range is described as between 0.2 to 20 meters. A complete Guidance module can be purchased for $999.00 (DJI, 2016e). 

   In summary, the DJI Guidance module offers a significant improvement in sense and avoid situations for obstacle clearance in small UAS. However, this is only part of the sense and avoid picture. The Federal Aviation Administration (FAA) is greatly concerned about small UAS in the national airspace and currently working with the National Air and Space Administration (NASA) to research a myriad of emerging technologies (FAA, 2009) that will best mitigate air traffic cooperation between manned and unmanned systems and eventually provide a complete sense and avoid capability for small UAS and provide a measure of comfort for manned aviation.

References:


ARM 2016 Cortex-A9 ProcessorARM. (2016). Cortex-A9 Processor. Retrieved from http://www.arm.com/cortex-a9.php 20160507133314770317554
Corrigan F 2015 Top Matrice 100 Programmable Sensor Drone With SDK ReviewCorrigan, F. (2015). Top Matrice 100 Programmable Sensor Drone With SDK Review. Retrieved from http://www.dronezon.com/drone-reviews/review-of-dji-matrice-100-sdk-drone-sensors-and-collision-avoidance-system/ 201605081033211218642473
Digital Trends 2016 Tech for the Way we LiveDigital Trends. (2016). Tech for the Way we Live. Retrieved from http://www.digitaltrends.com/about/ 201605071357581126976967
DJI 2016 3D SensingDJI. (2016a). 3D Sensing. Retrieved from http://www.dji.com/product/guidance/features#sub-feature 201605071321131122141481
DJI 2016 DJI StoreDJI. (2016b). DJI Store. Retrieved from http://store.dji.com/product/guidance?site=brandsite&from=buy_now_bottom#/spec 201605071419081274643183
DJI 2016 GuidanceDJI. (2016c). Guidance. Retrieved from http://www.dji.com/product/guidance 201605071309481873824358
DJI 2016 High-Precision Visual Positioning SystemDJI. (2016d). High-Precision Visual Positioning System. Retrieved from http://www.dji.com/product/guidance/features#sub-feature 20160507140558624188781
DJI 2016 Multi-sensor FusionDJI. (2016e). Multi-sensor Fusion. Retrieved from http://www.dji.com/product/guidance/features#sub-feature 201605071345421356029272
FAA 2009 Literature Review on Detect, Sense, and Avoid Technology for Unmanned Aircraft SystemsFAA. (2009). Literature Review on Detect, Sense, and Avoid Technology for Unmanned Aircraft Systems. Retrieved from http://www.tc.faa.gov/its/worldpac/techrpt/ar0841.pdf 201605071432561041083932
Prindle D 2015 DJI’s new Obstacle Avoidance Tech aims to make Drones Crash ProofPrindle, D. (2015). DJI’s new Obstacle Avoidance Tech aims to make Drones Crash Proof. Retrieved from http://www.digitaltrends.com/cool-tech/dji-obstacle-avoidance-matrice-100-guidance/ 201605071351471013692021