The fear that machines will ‘take over’ is one that is visceral in human consciousness, and more recently this fear tends, for some, towards drones.
Drones taking over the world is certainly an interesting cliché topic of discussion, but it is not the topic of this article. The topic of this article is a crushing problem at hand.
The security issues associated with drones and Unmanned Aerial Vehicles have become a reason of concern for many countries at both civilian and military capacities. Problems like military misuse of drones, potential conduction of terrorist activities with drones and privacy concerns associated with autonomous aircrafts are being addressed at different levels as the usage of Unmanned Aerial Vehicles (UAVs) is regulated in different states worldwide.
Although research concerning security fool-proofing of UAVs is being conducted all over the world, China has industrialized their drone market and they are actively involved in the construction of security systems for drones to ensure safe operation. One particular research paper titled Unmanned Optical Warning System For Drones published by Yin Zongdi, Song Qiudong, Han Guoqing and Zhu Meng; from China, presents a “photoelectric alarm system” that would identify a target and obtain legal evidence.
So why is it that researchers propose to use a “photoelectric alarm system” for ensuring security concerning autonomous aircrafts or UAVs?
This research paper describes an autonomous “optoelectronic warning system” targeted at drones that can simultaneously provide visible and infrared image information, and automatically identify the target along with the target’s characteristics or type.
The photovoltaic system is the front-end component of the anti-UAV system, so what it delivers is image information that presents the target situation most intuitively. In an unmanned intelligent system, the photoelectric alarm system is key to full automation.
How is it key to full automation, you may ask? In the absence of manned intervention, the interception measures of the security system will only be activated when the target is determined to be an incoming drone. Therefore, an unmanned anti-UAV photoelectric warning system affects the performance of the entire system directly.
Now, moving on to the technical part, how does this work exactly? Let us boil it down.
The system created by the Chinese researchers is equipped with a radar system or radar module. Obviously, the radar system is used to monitor drone activity. The radar transmits suspicious target information to the central controller.
What the central controller does immediately as it receives information from the radar is processing this information into meaningful data. In the case of drones, that meaningful data would be the exact, live location of the drone and its movements; helping with prediction of its future destinations over a time period.
This data from the central controller is taken to the optoelectronic system. Simply put, the job of the optoelectronic system is to identify and track illegally invaded targets by the rotation of the turntable on which the photoelectric source is mounted. Visible light field of view becomes smaller as the distance becomes longer, and can automatically focus to ensure market clarity.
Using a type of deep learning algorithm to identify the captured target, information relating to the identification of the target and tracking video image are transmitted from the optical server to the central server. If the identification type is a drone, the central controller controls the jammer that can in turn attack the drone.
After capturing the target, the visible light detector uses the convolutional neural network algorithm in deep learning to identify the target.
The infrared detector detects the small target, combining the detection results to identify and track the target in real time. Then, the recognition result and tracking videos are uploaded to the central controller.
The photoelectric system keeps on tracking a drone it is monitoring until it leaves the range of the system; as the system is limited by range, a number of modules would have to be installed for a wider range of security.
Citation: “Unmanned optical warning system for drones”, Zongdi Yin, Zongdi Yin, Qiudong Song, Qiudong Song, Guoqing Han, Guoqing Han, Meng Zhu, Meng Zhu, Proc. SPIE 10835, Global Intelligence Industry Conference (GIIC 2018), 108350Q (31 August 2018); doi: 10.1117/12.2503828; https://doi.org/10.1117/12.2503828