Safety Wing for Industry (SWI 2020) – An Advanced Unmanned Aerial Vehicle Design for Safety and Security Facility Management in Industries

T. Ananth Kumar; S. Arunmozhi Selvi, R.S. Rajesh, and C. Clorindal

Introduction

An increase in the use of innovative technology brings a security nightmare. These technologies as well have some backlogs to intrude. To overcome these backlogs, there are many practices specific to their application. External security is more focused, as it’s an essential issue for many multinational companies. Of course, many external forces are available for security, even though many unexpected attacks or vulnerabilities may happen.

The Quadcopters are usually loaded with multiple sensors and miniature processors. The Arduino, which is interfaced with IR sensors, will transmit the signal when an intruder trespasses the IR signal. The signal will be transmitted through the ZigBee module, which is also interfaced with the Arduino board. Arduino is open-source, which embeds both the software and hardware through small precised programming units. These units are enabled to design and manufacture microcontrollers in a single board and dynamic microcontroller kits for creating sensible interactive objects and control them in the real-world scenario both physically and digitally. ZigBee is a network protocol of type mesh, which is introduced to transport mini data chunks w'ith low power consumption for short distances. It uses an IEEE standard version 802.15.4 [l] commonly used in local area sensor data networks. Its transmitting distances to line-of-sight 10-100 meters using the 2.4 GHz ISM frequency band, are less costly than other wireless personal area networks (WPANs). This network works for many real-time applications like personal home automations and medical fields. The signal will be transmitted to the Raspberry Pi board, which w'ill be programmed to sense the signal when the intruder blocks the signal from the sensors, which are interfaced with the Arduino. The signal will contain the following details: a) Latitude and longitude of the sensors where they are placed; b) Approximate time of detection; c) Estimated Time Arrival of the quadcopter to arrive at the coordinates.

Industrial drones (Id) are known by multiple terms like Unmanned Aircraft System (UAV); Remotely Piloted Aircraft (RPA) is a standard vehicle which is used to monitor the terrains in and out of the Industrial Plant [2]. Usually, industrial plants are hazardous in nature, whatever may be the product it stores. The size of the plant may be small so as to accommodate only material where human intervention is not possible for regular operations. Areas like factories, refineries, power grids, cold mines, wastewater management, and even high nuclear or thermal plants are high risk for the human to work w'ith the process. More precautionary measures have to be taken to go through this system in the current scenario. Owing to this scenario, the Id enters into the situation, which can operate in any complex situations and even more challenging, to incorporate the surveillance of the industrial yard. This UAV is in the future to be connected w'ith the IoT applications to do wonders in the security features of the Industry, especially in the evolving civilian, military, and commercial use. The most commonly used drones are named Draganfly X6 [3] and Aerobatics UAV [4], specially tailored for the industrial and security sectors. This work deals with the design of unmanned aerial vehicles for surveillance in an industrial atmosphere for measuring a wide range. GPS based UAVs are used to observe the area where the unauthorized entry of a person or an intruder is spotted. This technology aims to achieve complete remote surveillance of an industry atmosphere where human reach couldn’t be possible.

How Drones Connect to the Industry

All commercial and military drones have a point-to-point connection with the wireless network at one end and the control system at the other end. An industrial drone communicates directly w'ith the industrial control system like mission control and data acquisition control. The drone is mounted w'ith many IoT sensors for the session control like temperature sensors, atmospheric pressure sensors, motion sensors, and humidity sensors. These sensors directly communicate w'ith the communication network to create bidirectional communication in the security aspect and logical protocol synchronization mapping. These drones are designed to capture aerial, thermal, and visual imagery, and even radiofrequency imagery of the factory units and subunits. There is some restriction in the drone imagery from nation to nation as its high-powered system can be hacked.

How Industries Like Energy and Utilities Use Drones for Industrial Applications

Due to the universal adoption and integration of all the automated markets like agriculture, mining, energy, oil and gas, emergency and life sciences, utilities and infrastructure need a unique novel UAV w'ith intelligence and compatible w'ith both hardware and software integration in a single knot. The IoT sensor’s measurements add spice to its novelty and eventually leads to great success. Thus, the whole automation industry and drone manufacturers found a solution that incorporates UAV with high-speed data analytics with IoT industrialization, cybersecurity, and drones w'ith the mounted sensor w'ithin [5]. There are standard guidelines to formulate all these intergradations into a procedure, and more practices have to have experimented. The existing scientists and engineers working currently on the various capabilities of the drones would work in the ground station, which will gather real-time information about the sensors from the mobile and fixed platforms.

Flight Mechanics and Control of Drones for Industrial Applications

When designing a drone, the application of which environment it's going to work in is one of the significant parts to be considered. Industrial, commercial, and military drones all work at different speeds and heights and also have their own performance characteristics. Thus, it is necessary to finalize the environment to work before designing the drone. The drone for the factory setting, the unit may be smaller with a much-closed space for movement, and the flight operation of RPA is designed for low fuel consumption. The battery backup has to be designed based upon the empirical values of the weight, power, and flight time. The weight of the drone is directly proportional to the battery life. The more weight consumed by the drone added with sensors, the more battery life is needed. Thus, altogether the weight of the sensor judges the battery life. The Standard Industrial Drone works from 20 minutes to several hours. A new latest innovative idea to move the drone beyond the visibility of the pilot sight is the BVLOS, the abbreviation for the Beyond-Visual-Line-Of-Sight. The approval for this setup is, as of now, not permitted. But many attempts are proved for the assurance of safety.

Automatons, authoritatively named as Unmanned Aerial Vehicles (UAV), are pilotless airplanes whose fundamental utility is to accumulate data on a domain and transfer it back to its controller. At the point when utilized in a mechanical background, automatons can shape a centerpiece of modern mechanization alongside IoT gadgets. Mechanical drones or UAVs are considered as the following breakaway in the progressing Industry 4.0 revolution. UAVs are mentioned as Quadcopter, which is a very useful tool for researchers of different fields of theory of flight control, realtime robotics systems. Most of the research in different universities has increased the performance of complex aerial imagery with the Quadcopter. A group of Quadcopters can fly in mid-air and do formations with many complex flying routines.

There are many reasons for choosing the Quadcopter as a testing vehicle as they are moderately low in cost, in more available sizes, and the simple mechanism of its working principle, which can be built and maintained by amateurs. Since multidisciplines like computer science, mechanical engineering, and electrical engineering are involved in modeling the Quadcopter, many academics from various fields combine to w'ork together to perform the Quadcopter operations.

Background Study

M.K. Khan et al. have approached the design and development of a quadcopter for surveillance purposes, which incorporates many sub-modules for communication and management [6]. It has been purposely designed to automatically control the Quadcopter by a flight controller APM2.7, which can be used for flight take-off and landing. The designed Quadcopter has its specification w'ith four brushless motors NTM 28-30s 800KV, and the Mobious mini action camera with Ю80 MP is mounted to capture the live video, and data are serially sent to the monitoring device for processing. The monitoring equipment is controlled by GPS Module N6N; a Ublox 6h

GPS is placed on the Quadcopter for the proper navigation. It uses an Orange Rx Т-Six 2.4GHz transmitter for transmission of the navigational position feedback to the controller.

Richard Steffen et al. have motivated in the real-time capable vision system of a UAV for the trajectory and surface revamping of the aerial images sequences [7]. The challenge was to provide a real-time vision with a single camera to present the design and methods for the same. The initialization process, map representation, the estimation process, exploration strategies, and loop closing detection were the various operations performed by the UAV. The critical point proposed was the Kalman-filter used for the estimation process, followed by the landmark-based map representation with a new initialization process and exploring strategy. Finally, contributing to a major impact on predicted camera trajectory and map accuracy. Landmark coordinates from a UAV without GPS was a complex challenge. The author added future works that can be applied with the sliding-window representation of the camera trajectory for better results in the approximation of motion in the Kalman-filter rather than linear prediction models which were represented by orientations and corresponding velocities of the actual camera.

Ankith Shetty et al. proposed a technique to control a drone by just using hand movements commonly known as hand gestures [8]. This paved the way for a human control interface to extract the ancient techniques of controlling using joysticks. This technique encourages more user-friendly human interaction with less knowledge of studying the instructions for controlling the different parameters of a drone. Unique algorithms are interfaced with different hand gesture input, which was feasible only by employing image processing techniques using the webcam. This was implemented with the various modules incorporated, such as hand gesture capture and gesture monitor, feature extraction, and pattern matching for recognition of the gesture captured. Thus, this paper influenced an approach to control a Quadcopter using gestures with the help of an Arduino and KK2.1.5 board.

Gaurav Choudhary et al. surveyed the current state-of-the-art IDS (intrusion detection system) mechanism, which focuses and deals with malicious attacks in the network [9]. When UAVs are networked, more attention has to be given to the security as well due to the malicious attacks over the air. This concern has developed to naturally derive and fix IDS vulnerabilities and attacks. The authors notably classify current IDS processes based on gathered sources, detection methods and comments, IDS recognition, and identification of intrusion. After the in-depth survey, the authors conclude with recent research challenges, observations, and potential directions to develop UAVIDS systems that meet the current trend's basic security function.

Sarra Berrahal et al. proposed a novel UAV for border surveillance solution in a Wireless Sensor Network to detect the intruders or trespassers [10]. The WSN communicates with the Quadcopters that were within the border of the network and detects network failures and maintenance. The Quadcopter, which was named VTail, automatically detects and locates the terrestrial sensors using RFID techniques and computes the position of the sensors to detect the intruders. This VTail detects coverage holes or failures and drops new sensors where failures occur at suitable positions, urgently relays data, captures the actual scenes, and automatically wakes up the ideal sensors in that zone to locate on the trajectory of an intruder.

Michael et al. suggested technology to automate an autonomous Quadcopter that incorporates a smartphone with Android OS, an Arduino board, creating an application that provides navigation instructions, preventing collisions and locating sensor data using Simultaneous Localization and Mapping (SLAM) [II]. As a result, the plan has an autonomous flight and navigation Quadcopter. Moreover, in the future, more progress in basic aerial control must be planned via commands in android application. For insufficient odometry and sensed data, reconsider an approach to SLAM. The Fast SLAM algorithm is effective in future mapping and position capabilities.

Carolyn Pearce et al. motivated that information acquiring is a vital part of military activities. UAVs have become a common tool for monitoring illegal activities, only in the outdoors as their size is huge and they rely more on GPS [12]. There are places where human intervention is not possible; such areas are where this UAV can be replaced as it is free from any harmful radiation in an unsafe zone. As a result, the authors proposed an indoor spatially aware autonomous UAV. It performs data transfer with the sensory input and IOIO Microcontroller to maintain interoperability. The Android OS is used for the Mission Control, and an Auto-Pilot System for the control of flying properties which is in par with the Department of Defense, MITRE, and academic partners. Thus, this work included a long group of engineering disciplines like computer science, electrical engineering, and systems engineering. SLAM technology in the latest edition provides on a national level security and other engineering futures.

Zeeshan Kaleem et al. have deployed an interesting work in UAV with interfaced sensors for various safety applications in the field of industries [13]. There are drones with uncontrolled deployment that may cause serious issues unexpectedly in more security-related areas like nuclear plants and VVIP house. The challenge, therefore, was to suggest a one to one architecture to support the erection successfully for Management Drone Monitoring (MDr) to catch Amateur Drone Monitoring (ADr). To do the above, the MDr should be able to detect, monitor, trap, and chase the ADr. As a result, the analysis of the current detection and tracking must be updated. Thus, while deploying the MDr it is necessary to check out the ADr’s behavior, which is a gap in the existing technology.

Alex G et al. developed a drone with an object notification controller using an onboard vision system with no localization of sensors and GPS [14]. A monocular computer vision algorithm is formulated, which can be applied in closed-loop control to find out the known object color. The information about height, range, and bearing of the aircraft is implemented with the corresponding feedback received for the object waypoints and the image frame size. A specific filter called a Kalman Filter was used for the removal of the noise generated in measuring the range from the object pixel. Thus, real-time object tracking of a low cost was implemented by the onboard Quadcopter using a standalone processor for the control of flight and image processing. The result expresses the settling time of under 10 s to 25 s for the stable closed-loop tracking analysis, with 10 for individual and 25 for coupled control.

Steven Rasmussen et al. described the UAV for the field test that worked together, which left the sensor unnoticed [15]. Also discussed were other normal operations including monitoring the area, detecting intruder vehicles, capturing the intruder’s image, and delivering its sources. This focuses on UAV, no sensors and perception algorithms, and just relays on off-board sensors. This work was experimented on a real-time environment where the UAV carried out all decision-making without human interference. Thus, as a final note, the authors experimented with an Automatic System encapsulating ground sensors, which were left unnoticed, where the UAV tracked intruders on their sensed network.

Alex Waller et al. have introduced a new' reactive UAV mobility planning approach to track high-risk areas [16]: PARCov Terminology of Autonomous Risk-sensitive Coverage Planner, consisting of a series of UAVs to monitor high-risk areas. PARCov is designed to conduct operations such as uninterrupted monitoring, providing good quality sensors, effective manipulation of sensor areas, and risk detection removal. Its efficiency was determined by simulating different risk models and increasing UAV numbers. The experiment's main UAV is the AscTec Pelican Quadcopter. As a result, it was motivated by the UAV’s viability to expand its automation services in the search and rescue mission of environmental monitoring, delivery of certain goods, and many active applications.

Rana et al.’s methodologies were to detect humans in isolated areas due to natural disasters. Most commonly, Quadcopters were used in areas like military, navigation, law enforcement, and aerial imaging [17]. The primary objective was to detect people from such natural disasters like flood, earthquake, and building destruction, where humans cannot intrude. This function was carried out using Raspberry Pi, and the processing of image activities was performed by the Open CV. Finally, the UAV captured the video streaming to assure the human detection and lively update of the control station to navigate. In order to fly longer, battery support was enabled, and a 3G dongle was supported for the large area surveillance.

Versatility of UAVs in Emergent Areas

In the early 2020s, a few companies such as Amazon and Facebook started utilizing the UAV, improving Internet access and delivery of products to the door; it is in the testing phase in some countries. In India, also, for overcoming the facts of COVID- 19, drones can be used for food delivery, grocery delivery, and also for medical emergency services [18]. Nowadays, commercial drones are getting cheaper, and industries have started designing application-specific drones, which are provoked to gather large amounts of visual data and for evolving business purposes. Most of the areas which started to use drones are agriculture, construction, surveying, mining of coal, entertainment, forestry, and for insurance claims. Automatons drones can perform assessments and reviews all the more rapidly, more inexpensively, and more securely than individuals or helicopters. Their advantage isn't in merely having the option to play out a similar undertaking all the more inexpensively, how'- ever altogether more now' and again also. By significantly expanding the measure of information assembled, new types of inspections are made conceivable. These are where drones have just begun to change enterprises.

Medical Care Control Drones

To battle COVID-19 in India, drones have emerged as technology partners in a few ways, like disinfectant spraying [19], medicine and grocery deliveries, temperature checking, surveillance, and authorities also are using drones to communicate messages and knowledge about lockdown measures. Tracing the animals in the forests additionally permits analysts to follow infections and diseases. The London School of Hygiene has utilized the UAVs with thermal imaging cameras and medicine to follow macaque developments within the territory of Palawan within the Philippines, a neighborhood where malaria may be a functioning danger. The capacity to follow these creatures gave further knowledge into the conceivable development of irresistible ailments and their bounces from creatures to people. During a comparable vein, Microsoft is utilizing ramble innovation to catch and test mosquitoes for irresistible infection. In a perfect world, this insight might be utilized to secure neighborhood occupants and, afterward, could help forestall pandemics before they begin [20].

Defense Services and Surveillance Drone Teams

The military drones, such as the Predator UAV, are much smaller in size with a longer flight time. The Government is spending a vast amount on developing drones for surveillance purposes exclusively as well as for offensive operations.

As the camera specification increases, various attributes in the camera have a great impact on drone technology. A thermal camera outfitted in the drone provides a quick response to identify victims, which is not possible with the naked eye. For example, Land Rover designed an autonomous drone with roof-mounted, thermal imaging operations. The drone has the ability to land directly on any moving vehicle on the road because of its integrated landing system. Drones have proved to be a valuable help during catastrophic events. In the fallout of tropical storms and quakes, UAVs have been utilized to find the persons in the natural disaster and meet their casualty. The AI drones have the capability of taking eight images per object. Neurula's algorithm-based drone is trained by deep neural networks and is useful in inspecting cell towers, wind turbines, pipelines, and electrical infrastructure in the industries [21].

Forest and Weather Forecasting Drones

Moderates achieve creative methods to help combat this pattern and ensure and take global biological processes into consideration. Along with the geospatial representation, UAV's are now used to screen and track animals.

The Zack Clark, Drone Integration Specialist at America investigated 300 miles of pipeline or 500 miles of the utility line. A specialist isn't out there to educate the automaton administrator on anything. The automatons are out of range to take care of continuous video back, and they are out of the view for administrators. Automatons gather tremendous measures of information, and this information should be looked into and handled later to distinguish absconds in the pipelines.” Most research is equipped with different programming standard for acquiring information about the study of atmospheric climate. For example, in the Saildrones pilot project, the company allotted six drones which collect the data continuously from the shore to make the weather forecasts accurate. Global waste management has been increased due to recycling and biodegradation. Nowadays, automated drones are used to clean the oceans.

Energy and Mining Industrial Drones

Though mainstream energy is available, petroleum products, despite everything, remain a key vitality hotspot for the world. The structures for removal, refining and transport of oil and gas are a significant component of the organisation. These are required to maintain compliance with guidelines and regulations on a daily basis. When unmanned drones are used, inspection work on the remote should be possible safely. Using the thermal sensors, a few drones can very easily detect quicker than human detection. Another zone, where automatons have shown promise, is in setting up new destinations for the creation of energy in industries.

Reserves of metal or rock or minerals are hard to gauge. In any case, with a special type camera, rambles can catch volumetric information on reserves and review mining activities from the air. The application-specific drones are utilized by digging organizations for estimating materials, reviewing activities, and boosting security. The framework is completely self-governing and put away in on-location lodging that can autonomically trade cameras and batteries.

Construction Planning and Infrastructure Development Drones

The emerging software developers have created a solution to analyze construction activities, and their progress of work in the real-time video captured regularly. As the surveying in civil engineering is still a critical issue in construction and planning, the drone makes it simple and most important. Even the soil type throughout the construction can be monitored effectively and progressively. The high building can also be monitored effectively from the ground station. There is special software available on subscription bases that can be paired with one or more UAVs for complete monitoring and site management. This, in the future, could also serve a more useful purpose in physical infrastructure like lifting polymer bricks.

Urban planning is a major task for metropolitan cities, which requires complete knowledge of the plan in certain areas, which leads to a whole idea of dense population area. Based upon the drone images and videos, the municipal corporation can end up with an idea of how to accommodate the green space. This reduces the workmanship of humans and time for evaluation.

Insurance Claims and Retail Operations in Industries

For all the insurance cases, previously inspectors and technical people used to inspect the property of insurance for the damage. This can be monitored by the drone, which can provide a more detailed assessment and clear pictures of the asset. Most drone manufacturers focused on the insurance industries, which may lead to claims based upon the roof inspection made in many industrial places.

Drones are used as a transporter in some areas, but this has not seen a great deal of success. They deliver goods in places where humans move or report. Amazon was the first to deliver through drones in 2016.

Internet and Communication Trends in Industries

Telecommunication towers that are at high risk in maintenance can be monitored and inspected frequently by the drones, which reduces the radiation affecting humans. They efficiently assess the damage to help the maintenance team proceed further, which is the most time-consuming factor manually. Aquila, a solar-powered drone, was introduced to provide global Internet access for the rural areas experimented by Facebook. Google followed by introducing a solar-powered drone named Project Loon designed by the Titan aerospace.

Tourism- and Hospitality-Based Drones in Industries

Drones, in their advanced stage, are used as a recreational tool outdoors, and are responsible for many photos in the outdoors from aerial landscapes to sports footage. Many filmmakers use this drone to take breathtaking photos and award-winning films. Sense Fly is one among them to take unbelievable shots that people can't imagine. Traditional hotels make use of the drone to travel to a new location on demand. They even provide hospitality to the guest by serving them with their needs.

Sports and Live Entertainment

There are drones used for entertainment, such as event photography and film making. Drones can directly entertain in the sky named as Disney. Drawing patterns in the sky and puppeteers have all been an entertainment giant. Drones are becoming increasingly common in outdoor and indoor sports. They have added more flavor to the televised sports. Nowadays, Skycam is a necessity for any professional arena sports broadcast.

Security, Underground Economies and Crime-Fighting Drones

People misuse drones for illegal purposes, where there are certain limitations to enter. Even drug transportation is also a major issue. Many military borders have been victims to these types of drones. Special security has been given by these drones for both commercial and personal use. Many companies are using drones for surveillance and detection for industrial, commercial, and residential properties. Especially high-security affluent areas are secured by the batch of drones to minimize the operational cost with the human.