1. An unmanned aerial vehicle (UAV) system for obstacle mapping and avoidance, comprising: a UAV equipped with the proximity sensors, which generates the sensory data; a two-neuron perpetual system network that processes the generated sensory data into the motor commands; a correlation-based learning mechanism with a synaptic scaling for adjusting the synaptic weights within the network; and an Android application for displaying the generated sensory data and dynamically creating an obstacle map. 2.The UAV system of claim 2, wherein the system is configured to adapt neuronal dynamics of the network based on different physical phenomena effects. 3.The UAV system of claim 2, wherein the system utilizes various angles formed by significant functions of the network to navigate intersection locations and dead ends. 4.The UAV system of claim 2, configured to mimic autonomous drone behaviour and obstacle avoidance strategies using a modified version of the V-REP Simulator. 5.The UAV system of claim 2, wherein the two-neuron perpetual system network applies post-processing neural units for pitch and yaw management to convert sensory information into motor commands. 6.The UAV system of claim 2, wherein the system's synaptic weights increase or decrease based on the perceived effectiveness of the UAV's perception of its surroundings, leading to adaptive neuronal dynamics. 7.The UAV system of claim 2, wherein the UAV is configured to inspect and navigate within a turbulent, unstructured environment by avoiding obstacles, sneaking around corners, and not becoming stuck. 8.The UAV system of claim 2, configured to adapt the synaptic weights within the two-neuron perpetual system network based on the UAV's perception of its surroundings, thereby enhancing the UAV's ability to recognize and avoid obstacles. 9.The UAV system of claim 2, wherein the varying angles formed by the network's significant functions enable the UAV to handle intersection locations and dead ends without becoming stuck. 10.A method for generating an obstacle map using an unmanned aerial vehicle (UAV), the method comprising: deploying a UAV equipped with the proximity sensors to the detect physical objects and their proximity; processing the sensory data through a synaptic plasticity-based repeating network, specifically a two-neuron perpetual system network, to derive motor commands; adjusting the synaptic weights within the network based on a correlation-based learning mechanism and the synaptic scaling; directing the UAV's pitch and yaw movements based on the motor commands to navigate and avoid the obstacles in the region; and displaying the UAV's sensory data on an Android application to dynamically create an obstacle map. UNMANNED AERIAL VEHICLE (UAV) SYSTEM FOR OBSTACLE MAPPING AND AVOIDANCE Abstract This invention discloses a unique system that utilizes an unmanned aerial vehicle (UAV) equipped with proximity sensors for autonomous obstacle mapping and avoidance. The UAV processes sensory data through a two-neuron perpetual system network, transforming it into motor commands. The network employs a correlation-based learning mechanism and synaptic scaling to adjust synaptic weights, thereby enhancing navigation and obstacle recognition capabilities. Moreover, the system integrates an Android application to display the UAV's sensory data, facilitating real-time mapping of the environment. The UAV system can adapt neuronal dynamics based on different physical phenomena, uses the network's functions for efficient navigation, and mimics autonomous drone behavior via a modified V-REP Simulator. The invention also discloses a method for generating obstacle maps using the UAV system, promising a robust solution for navigating turbulent, unstructured environments and providing dynamic obstacle mapping. , Claims:Claims :

1. An unmanned aerial vehicle (UAV) system for obstacle mapping and avoidance, comprising: a UAV equipped with the proximity sensors, which generates the sensory data; a two-neuron perpetual system network that processes the generated sensory data into the motor commands; a correlation-based learning mechanism with a synaptic scaling for adjusting the synaptic weights within the network; and an Android application for displaying the generated sensory data and dynamically creating an obstacle map. 2.The UAV system of claim 2, wherein the system is configured to adapt neuronal dynamics of the network based on different physical phenomena effects. 3.The UAV system of claim 2, wherein the system utilizes various angles formed by significant functions of the network to navigate intersection locations and dead ends. 4.The UAV system of claim 2, configured to mimic autonomous drone behaviour and obstacle avoidance strategies using a modified version of the V-REP Simulator. 5.The UAV system of claim 2, wherein the two-neuron perpetual system network applies post-processing neural units for pitch and yaw management to convert sensory information into motor commands. 6.The UAV system of claim 2, wherein the system's synaptic weights increase or decrease based on the perceived effectiveness of the UAV's perception of its surroundings, leading to adaptive neuronal dynamics. 7.The UAV system of claim 2, wherein the UAV is configured to inspect and navigate within a turbulent, unstructured environment by avoiding obstacles, sneaking around corners, and not becoming stuck. 8.The UAV system of claim 2, configured to adapt the synaptic weights within the two-neuron perpetual system network based on the UAV's perception of its surroundings, thereby enhancing the UAV's ability to recognize and avoid obstacles. 9.The UAV system of claim 2, wherein the varying angles formed by the network's significant functions enable the UAV to handle intersection locations and dead ends without becoming stuck. 10.A method for generating an obstacle map using an unmanned aerial vehicle (UAV), the method comprising: deploying a UAV equipped with the proximity sensors to the detect physical objects and their proximity; processing the sensory data through a synaptic plasticity-based repeating network, specifically a two-neuron perpetual system network, to derive motor commands; adjusting the synaptic weights within the network based on a correlation-based learning mechanism and the synaptic scaling; directing the UAV's pitch and yaw movements based on the motor commands to navigate and avoid the obstacles in the region; and displaying the UAV's sensory data on an Android application to dynamically create an obstacle map.

Description:UNMANNED AERIAL VEHICLE (UAV) SYSTEM FOR OBSTACLE MAPPING AND AVOIDANCE
Field of the Invention
[0001] This invention pertains to the field of drone technology and artificial intelligence. It specifically relates to an unmanned aerial vehicle (UAV) system that uses a synaptic plasticity-based perpetual system network for autonomous obstacle mapping and avoidance, with an accompanying Android application for displaying real-time sensory data and mapping.
Background
[0002] The background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.
[0003] The field of robotics and automation has seen significant advancement in the last few decades, becoming an integral part of various industries. In particular, unmanned aerial vehicles (UAVs), popularly known as drones, have emerged as highly versatile tools. Their applications range from aerial photography and film-making to logistics and delivery, search and rescue operations, environmental monitoring, and more. This invention pertains to the use of automated drones for generating a map of an unknown area, a domain of growing interest given the potential such technology holds for surveying and exploration in remote or inaccessible regions.
[0004] Traditionally, map creation and surveying were time-consuming and labor-intensive processes. They often required human surveyors to physically traverse the landscape, posing numerous challenges including risk to human life, high costs, and the difficulty of accessing some terrains. Satellite imaging and global positioning systems (GPS) brought about revolutionary changes, offering the ability to conduct wide-scale surveys from space. However, these methods still have limitations in terms of resolution, cloud cover interference, and the inability to capture indoor or underground spaces effectively.
[0005] On the other hand, drones, with their ability to fly at low altitudes and navigate narrow spaces, can overcome many of these limitations. Initially, drones were manually operated, requiring skilled pilots to navigate them accurately. The advent of automation and artificial intelligence (AI) technology has dramatically changed this landscape. Drones can now be programmed to fly specific routes, avoid obstacles, and perform complex tasks, reducing the need for human involvement.
[0006] The advent of automated drones opens the door to a multitude of new applications, among which is the exploration and mapping of unknown territories. However, the challenge lies in developing reliable, robust, and efficient algorithms that enable drones to autonomously navigate these areas while gathering necessary data to build accurate maps. Moreover, these drones must also be capable of dealing with unpredictable elements in unknown environments such as unexpected obstacles, changes in weather conditions, and varying light conditions, amongst others.
[0007] Existing research in this area has resulted in semi-autonomous drone systems that can map known environments. These systems typically use predefined flight paths and are largely dependent on GPS for navigation. This method, while effective in certain circumstances, is not suitable for mapping unknown areas, particularly those where GPS signals are unreliable or non-existent. Furthermore, these systems often lack the ability to dynamically adapt to changes in the environment or make real-time decisions, essential factors for autonomous exploration.
[0008] There is thus a clear need for a drone system capable of fully autonomous navigation and mapping in unknown environments. This invention proposes a drone system that uses advanced AI algorithms for real-time decision-making and mapping. The system relies on onboard sensors, rather than GPS, to gather data and determine its position, making it suitable for use in GPS-denied environments. Additionally, the drone is capable of dynamically adapting its flight path based on the data gathered, enabling it to avoid obstacles and efficiently explore the environment.
[0009] In essence, the technology detailed in this invention is set to usher in a new era in the field of autonomous drones and robotic mapping, demonstrating a significant leap from previous technologies. The potential applications of this technology are vast, ranging from search and rescue operations in disaster-stricken areas, archaeological explorations, environmental studies, infrastructure inspections, to military reconnaissance and beyond. The ongoing advancements in AI and drone technology, as encapsulated in this invention, signify a promising future for the automation industry and its multifaceted applications..
[00010] All references, including publications, patent applications, and patents, cited herein are hereby incorporated by reference to the same extent as if each reference were individually and specifically indicated to be incorporated by reference and were set forth in its entirety herein.
[00011] It also shall be noted that as used herein and in the appended claims, the singular forms “a”, “an”, and “the” include plural referents unless the context clearly dictates otherwise. This invention can be achieved by means of hardware including several different elements or by means of a suitably programmed computer. In the unit claims that list several means, several ones among these means can be specifically embodied in the same hardware item. The use of such words as first, second, third does not represent any order, which can be simply explained as names.
Summary
[00012] The following presents a simplified summary of various aspects of this disclosure in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements nor delineate the scope of such aspects. Its purpose is to present some concepts of this disclosure in a simplified form as a prelude to the more detailed description that is presented later.
[00013] The following paragraphs provide additional support for the claims of the subject application.
[00014] This invention pertains to the field of drone technology and artificial intelligence. It specifically relates to an unmanned aerial vehicle (UAV) system that uses a synaptic plasticity-based perpetual system network for autonomous obstacle mapping and avoidance, with an accompanying Android application for displaying real-time sensory data and mapping.
[00015] This invention delves into a system and methodology employing an unmanned aerial vehicle (UAV), equipped with proximity sensors for autonomous obstacle detection, mapping, and avoidance. The UAV works in tandem with a two-neuron perpetual system network, which processes sensory data, converting it into motor commands that guide the UAV's movements. The system utilizes a correlation-based learning mechanism alongside synaptic scaling, enabling the adjustment of synaptic weights within the network. Additionally, the system includes an Android application, presenting the UAV's sensory data to the user and facilitating the dynamic creation of an obstacle map.
[00016] The innovation lies in the two-neuron perpetual system network's ability to adapt neuronal dynamics based on the effects of different physical phenomena. This feature enhances the system's capacity to respond to changing environments and unpredicted obstacles, contributing significantly to the safety and efficiency of the UAV's navigation. Additionally, the system utilizes various angles formed by significant functions of the network, enabling the UAV to handle intersection locations and dead ends effectively. This provides an added layer of navigation proficiency, ensuring that the UAV doesn't become stuck in complex environments.
[00017] The system mimics autonomous drone behavior using a modified version of the V-REP Simulator, a virtual robot experimentation platform. This feature allows for a detailed testing and tuning phase, enabling the refinement of the drone's autonomous behavior before deploying it in real-world situations. Furthermore, the two-neuron perpetual system network employs post-processing neural units for managing the UAV's pitch and yaw. The sensory information conversion into motor commands, again, serves to increase the UAV's maneuverability and obstacle avoidance capabilities.
[00018] A noteworthy aspect of the invention is its adaptive nature. The synaptic weights of the system increase or decrease based on the UAV's perception of its surroundings. This dynamic response, dubbed 'adaptive neuronal dynamics,' allows for a fluid adjustment to changes in the environment, ensuring optimal performance at all times.
[00019] Specifically designed to operate in turbulent, unstructured environments, the UAV is capable of avoiding obstacles, sneaking around corners, and preventing itself from becoming stuck. This gives the UAV a higher degree of environmental adaptability, making it ideal for mapping and exploring complex, unpredictable landscapes.
[00020] The invention also outlines a method for generating an obstacle map using the described UAV. This process involves deploying the UAV, which utilizes its sensors to detect physical objects and their proximity. The sensory data is then processed through the two-neuron perpetual system network to derive motor commands. The synaptic weights within the network are adjusted using a correlation-based learning mechanism and synaptic scaling. This information then directs the UAV's pitch and yaw movements to navigate and avoid obstacles in the area. The UAV's sensory data is displayed on an Android application, dynamically creating an obstacle map.
[00021] The V-REP Simulator can be used to simulate the two-neuron perpetual system network and the UAV's behavior. This feature allows for the virtual testing of the system and any necessary adjustments before deploying the UAV in real-world scenarios. The synaptic weights within the network adjust based on the UAV's perception of its surroundings, resulting in the continuous alteration and adaptation of the neuronal dynamics of the network.
[00022] The UAV uses various angles formed by the significant functions of the network to handle intersection locations and dead ends. This process ensures continuous navigation without getting stuck, making the UAV capable of autonomously navigating complex environments. The Android application processes the visual feed, the proximity sensor's trigger, the gap between objects and the sensor, the location of the UAV, and its orientation to create a dynamic obstacle map. This feature provides a real-time visual representation of the UAV's environment, aiding users in understanding the area being explored.
[00023] The invention concludes with a discussion on future implementations of the method. The inventors envision deploying actual UAVs or multiple UAVs in a real environment to observe adaptive obstacle avoidance behavior. This focus on real-world deployment underlines the inventors' confidence in the system's readiness and practical utility in a variety of scenarios, from search and rescue to environmental surveying and beyond.
[00024] In essence, this invention represents a significant advancement in autonomous UAV navigation and mapping. It introduces a system that not only efficiently avoids obstacles but also learns and adapts to its environment. The dynamic creation of an obstacle map provides a user-friendly interface for understanding the UAV's operations, making this technology a promising tool for a broad spectrum of applications..
Brief Description of the Drawings
[00025] The features and advantages of the present disclosure would be more clearly understood from the following description taken in conjunction with the accompanying drawings in which:
[00026] Fig. 1 illustrates an unmanned aerial vehicle (UAV) system 100 designed to perform obstacle mapping and avoidance, in accordance with an embodiment of the present disclousre.
[00027] Fig. 2 illustrates a method 200 for creating an obstacle map using a UAV, in accordance with an embodiment of the present disclosure.
[00028] Fig. 3 illustrates a network overview of adaptive neural control, in accordance with an embodiment of the present disclosure.
[00029] Fig. 4 illustrates a neural control arrangement system, in accordance with an embodiment of the present disclosure.
[00030] Fig. 5a and 5b illustrates a Client-server architecture of web-services and a RESTful communications between user and UAV/drone, respectively, in accordance with an embodiment of the present disclosure.
[00031] Fig. 6 illustrates simulation of obstacles map, in accordance with an embodiment of the present disclosure.
[00032] Fig. 7 illustrates an Android applications showing visual feed from the UAV/drone, in accordance with an embodiment of the present disclosure.
Detailed Description
[00033] In the following detailed description of the invention, reference is made to the accompanying drawings that form a part hereof, and in which is shown, by way of illustration, specific embodiments in which the invention may be practiced. In the drawings, like numerals describe substantially similar components throughout the several views. These embodiments are described in sufficient detail to claim those skilled in the art to practice the invention. Other embodiments may be utilized and structural, logical, and electrical changes may be made without departing from the scope of the present invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is defined only by the appended claims and equivalents thereof.
[00034] The use of the terms “a” and “an” and “the” and “at least one” and similar referents in the context of describing the invention (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. The use of the term “at least one” followed by a list of one or more items (for example, “at least one of A and B”) is to be construed to mean one item selected from the listed items (A or B) or any combination of two or more of the listed items (A and B), unless otherwise indicated herein or clearly contradicted by context. The terms “comprising,” “having,” “including,” and “containing” are to be construed as open-ended terms (i.e., meaning “including, but not limited to,”) unless otherwise noted. Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein. All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g., “such as”) provided herein, is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the invention.
[00035] This invention pertains to the field of drone technology and artificial intelligence. It specifically relates to an unmanned aerial vehicle (UAV) system that uses a synaptic plasticity-based perpetual system network for autonomous obstacle mapping and avoidance, with an accompanying Android application for displaying real-time sensory data and mapping.
[00036] Pursuant to the "Detailed Description" section herein, whenever an element is explicitly associated with a specific numeral for the first time, such association shall be deemed consistent and applicable throughout the entirety of the "Detailed Description" section, unless otherwise expressly stated or contradicted by the context.
[00037] Fig. 1 illustrates an unmanned aerial vehicle (UAV) system 100 designed to perform obstacle mapping and avoidance, in accordance with an embodiment of the present disclousre. The UAV system 100 includes a UAV equipped with proximity sensors 102, a two-neuron perpetual system network 104, a correlation-based learning mechanism 106 with synaptic scaling, and an Android application 108.
[00038] In an embodiment, the UAV serves as the system's main hardware element, designed to navigate autonomously in diverse environments. Equipped with a range of proximity sensors, the UAV is capable of detecting physical objects and discerning their proximity in real-time. These sensors, which could include radar, LiDAR, ultrasonic, or infrared, continually scan the surrounding environment to identify potential obstacles. The real-time data they produce encapsulates the UAV's situational context, providing details about the physical landscape, object dimensions, distances, and any other elements that might obstruct the UAV's flight path. This continuous flow of sensory data is crucial, enabling the UAV to assess and adapt to its environment, facilitate real-time navigation decisions, and avoid collisions. Thus, the UAV can safely and effectively conduct operations, even in undiscovered or dynamically changing regions.
[00039] In an embodiment, the two-neuron perpetual system network is central to the UAV system's operations. As the UAV's sensors gather environmental data, which is relayed to the network for processing. The network's primary role is to translate the complex sensory data into motor commands, acting as a bridge between the UAV's perception of its surroundings and its responsive movements. The network structure, despite its minimalistic design, proves to be a potent computational model that utilizes fundamental principles of neural networks, which are known for their ability to process vast amounts of data efficiently and adaptively. Mimicking the functionality of a human brain on a simpler scale, the two-neuron perpetual system network effectively interprets environmental cues and generates appropriate responses. These responses, or motor commands, guide the UAV's navigation, enabling the UAV system to traverse its environment, avoid obstacles, and adjust its course as necessary, all in real-time, showcasing the effectiveness and power of such a compact network design.
[00040] In an embodiment, alongside the two-neuron perpetual system network, the UAV system utilizes a correlation-based learning mechanism with synaptic scaling. The learning mechanism takes inspiration from the principle of synaptic plasticity found in biological systems, wherein synaptic strength, or weights, dynamically change based on the neuron's activity level. Within the UAV system, these synaptic weights are adjusted based on how effectively the UAV interacts with its environment. For example, if the UAV successfully navigates around an obstacle, the corresponding synaptic weights might be increased. Conversely, if the UAV nearly collides with an object, the related weights might be reduced. Such adaptive adjustment of synaptic weights thus impacts the output of the two-neuron network, directly influencing the motor commands sent to the UAV. Consequently, the UAV's ability to recognize and avoid obstacles continuously improves, making the mechanism a powerful tool for fostering machine learning and autonomy in complex, real-world environments.
[00041] In an embodiment, the UAV system incorporates an Android application, functioning as a user-friendly interface that presents essential data gathered by the UAV in an accessible and interpretable manner. The application displays a range of the UAV's sensory data, including visual feeds captured by on-board cameras, triggers from proximity sensors, distance between the UAV and detected objects, as well as the UAV's location and orientation. Such information provides the user with a comprehensive, real-time overview of the UAV's operations. Moreover, a significant feature of the application is its ability to dynamically create an obstacle map. Utilizing the sensory data provided by the UAV, the application generates a visual map outlining the detected obstacles in the UAV's environment. The obstacle map continually updates as the UAV navigates, providing a visual aid for understanding the UAV's interactions with its surroundings. Thus, this application enhances user engagement and understanding, facilitating a more effective and informative UAV operation experience.
[00042] Upon deployment, the UAV embarks on its mission to survey an area, mapping obstacles and charting a path through its environment. Its proximity sensors - integral components of the UAV - are continually active, scanning the surroundings and detecting physical objects that could obstruct its path. The real-time sensory data is crucial for the UAV to navigate safely and is continuously fed into the two-neuron perpetual system network for processing.
[00043] Comprising two interconnected neural units, the two-neuron perpetual system network operates as the UAV system's computational core that interprets the incoming sensory data and uses the sensory data to determine the UAV's motor commands. These commands could involve changes in direction, altitude adjustments, speed modifications, and more, essentially controlling the UAV's navigation.
[00044] Intrinsic to the network's design is its dynamic and adaptive nature. It continually adjusts to changing conditions, mirroring the concept of synaptic plasticity in biological systems. As the UAV encounters new obstacles or navigates through diverse environments, the network adapts, adjusting its computational processes based on these experiences. Such adaptive quality allows the UAV to efficiently react to unforeseen obstacles or changes in its environment, reinforcing its autonomous capability, and ensuring a robust performance even in unpredictable scenarios.
[00045] The correlation-based learning mechanism, augmented with synaptic scaling, is fundamental to the UAV system's adaptability and intelligence. As the UAV interacts with its surroundings and navigates through varied environments, the learning mechanism adjusts the network's synaptic weights based on the effectiveness of those interactions. Correlation-based learning, a variant of Hebbian learning - often summarized by the phrase "neurons that fire together, wire together" - is the underlying principle of the learning mechanism. In the system, changes in synaptic weights are proportional to the correlation between activities of pre- and post-synaptic neurons. Essentially, if a neuron's input consistently leads to the activation of another neuron, the synaptic strength between these neurons increases. The learning mechanism allows the UAV system to 'learn' from successful interactions and refine its obstacle detection and avoidance strategies accordingly. Complemented with synaptic scaling, the learning mechanism ensures that the synaptic weights are kept within manageable limits, preventing run-away excitation or inhibition within the network, which thus balances the need for adaptability with system stability, allowing the UAV to efficiently navigate complex environments.
[00046] In an embodiment, the Android application serves as the human-machine interface of the UAV system. The application receives real-time sensory data from the UAV, displaying it in an easily digestible format. Furthermore, the application generates a dynamic obstacle map based on the UAV's location and detected objects. The obstacle map is continually updated as the UAV navigates its surroundings, providing a real-time representation of the environment.
[00047] A practical application of the system might involve deploying the UAV in an unknown or hostile environment. The UAV, guided by its sensors and the two-neuron perpetual system network, would navigate the environment, avoiding obstacles and adapting to changes in its surroundings. The system's adaptability, facilitated by the correlation-based learning mechanism, allows the UAV to learn from its interactions with the environment and improve its navigation capabilities over time.
[00048] Simultaneously, the Android application provides the operators with a real-time obstacle map, allowing them to understand the UAV's surroundings without being physically present. Such functionality makes the UAV system a valuable tool in situations where human presence might be dangerous or impractical, such as disaster response or military reconnaissance.
[00049] Future implementations of the system could involve integrating more advanced sensory equipment into the UAV or expanding the capabilities of the two-neuron perpetual system network. For instance, the UAV could be equipped with LiDAR or infrared sensors to improve obstacle detection in

challenging environmental conditions. Similarly, the neural network could be expanded to include more neurons or layers, enhancing the system's computational power and adaptability.
[00050] Additionally, multiple UAVs could be used in a coordinated manner to survey larger areas more quickly. Each UAV would generate its own sensory data and obstacle map, which could then be combined into a comprehensive representation of the larger environment. The multi-UAV approach could further increase the system's effectiveness in various applications, from search and rescue operations to environmental monitoring.
[00051] In this embodiment, the UAV system is capable of adapting the neuronal dynamics of the two-neuron perpetual system network based on different physical phenomena. For instance, the physical phenomenon could be variations in weather conditions, terrain, or physical obstructions in the environment. The system's synaptic weights, which are adjusted in real-time through the correlation-based learning mechanism, change as a response to these varying phenomena. Such adaptation allows the UAV to better understand and react to its surroundings, enhancing its navigational efficiency and obstacle avoidance capabilities.
[00052] This embodiment focuses on the UAV system's ability to utilize various angles formed by the significant functions of the two-neuron perpetual system network for navigation. These angles enable the UAV to effectively navigate intersection locations and dead ends without becoming stuck. Depending on the computed angle, the UAV might turn, ascend, descend, or hover to avoid obstacles.
[00053] Here, the UAV system's ability to mimic autonomous drone behavior using a modified version of the V-REP Simulator is highlighted. The V-REP simulator provides a virtual environment for the UAV, allowing it to practice obstacle avoidance and navigation strategies without real-world risks. Such training can help improve the UAV's operational efficiency and safety in the actual environment.
[00054] In this embodiment, the two-neuron perpetual system network's role in post-processing neural units for pitch and yaw management is emphasized. The sensory information inputted to the system is processed and converted into motor commands for the UAV. The network's configuration helps ensure the UAV maintains its balance and control while executing its navigation commands, contributing to successful obstacle avoidance.
[00055] In an embodiment, the UAV system's synaptic weights increase or decrease based on the perceived effectiveness of the UAV's perception of its surroundings, leading to adaptive neuronal dynamics. The system is configured to adapt the synaptic weights within the two-neuron perpetual system network based on the UAV's perception of its surroundings.
[00056] These embodiments highlight the adaptive nature of the UAV system's synaptic weights based on its perception of the surroundings. If the UAV perceives a successful interaction with its environment (e.g., an obstacle is effectively avoided), the synaptic weights are increased. Conversely, if the UAV perceives a less successful interaction (e.g., the UAV gets too close to an obstacle), the synaptic weights are decreased. Such adaptability of synaptic weights contributes to the UAV's ability to learn from its environment and improve its navigational abilities over time.
[00057] In an embodiment, the UAV system is configured to inspect and navigate within a turbulent, unstructured environment by avoiding obstacles, sneaking around corners, and not becoming stuck. The varying angles formed by the network's significant functions enable the UAV to handle intersection locations and dead ends without becoming stuck.
[00058] These claims detail the UAV's capability to navigate within a turbulent, unstructured environment. By utilizing various angles formed by the network's significant functions, the UAV can effectively avoid obstacles, maneuver around corners, and navigate intersections and dead ends without becoming stuck. Such feature empowers the UAV to function effectively even in challenging or unfamiliar environments.
[00059] Fig. 2 illustrates a method 200 for creating an obstacle map using a UAV, in accordance with an embodiment of the present disclosure. The key steps of the method 200 include the deployment of a sensor-equipped UAV, processing of the sensory data, adjustment of synaptic weights, UAV movement direction, and data display via an Android application. Here is a more detailed description of each step; At step 202, a UAV equipped with sensors is deployed to a designated area. The sensors, which may include radar, LiDAR, or other proximity sensors, are crucial for detecting physical objects and determining their proximity. As the UAV traverses the environment, the sensors continuously scan the surroundings, generating data about the location and distance of any detected obstacles. At step 204, the sensory data gathered by the UAV is fed into a synaptic plasticity-based repeating network, more specifically, a two-neuron perpetual system network that uses the principle of synaptic plasticity, where changes in neural activity can lead to changes in the strength of synaptic connections. The network processes the sensory data, transforming it into actionable motor commands that will guide the UAV's navigation. At step 206, the synaptic weights within the two-neuron perpetual system network are adjusted based on a correlation-based learning mechanism and synaptic scaling. Essentially, the effectiveness of the UAV's perception of its surroundings influences the adjustment of these weights. If the UAV effectively identifies and avoids an obstacle, the corresponding synaptic weights are increased. Conversely, if the UAV fails to avoid an obstacle effectively, the synaptic weights are decreased. Such dynamic adjustment enables the UAV to learn from its environment and enhance its navigational abilities over time. At step 208, once the motor commands are derived, they are used to direct the UAV's movements, specifically its pitch and yaw. Pitch and yaw control allows the UAV to ascend or descend, and turn left or right, respectively. These movements enable the UAV to navigate around obstacles, ensuring that it can traverse its environment safely and efficiently. At step 210, the UAV's sensory data is displayed on an Android application that provides a user-friendly interface, showing real-time data about the UAV's location, orientation, detected obstacles, and their proximity. Moreover, the application dynamically generates an obstacle map based on the data received from the UAV. The obstacle map offers an easily interpretable visualization of the UAV's environment, enabling operators to understand the landscape and potential obstacles that the UAV encounters.
[00060] In another embodiment, a simulation environment, specifically a modified version of the V-REP Simulator, is used to test and train the two-neuron perpetual system network and the UAV's behavior. By virtually reproducing potential real-world scenarios, the UAV can learn and adapt its behavior safely and effectively before deployment in the actual environment. The simulator allows for continuous fine-tuning and validation of the network and UAV's responses, ensuring optimal performance when put into real-world use.
[00061] This aspect highlights the adaptive nature of the two-neuron perpetual system network. The network's synaptic weights are adjusted based on the UAV's perception of its environment. If the UAV effectively avoids an obstacle, the corresponding synaptic weights are increased. If the UAV fails to effectively avoid an obstacle, the weights are decreased. The adaptive mechanism allows the UAV to learn from its past actions, thus improving its navigation capabilities over time.
[00062] This embodiment describes how the UAV uses different angles formed by the significant functions of the two-neuron perpetual system network to navigate complex environments. Such feature enables the UAV to handle intersections and dead ends effectively without becoming stuck, ensuring smooth, continuous navigation.
[00063] This embodiment emphasizes the Android application's role in processing the UAV's sensory data to create a dynamic obstacle map. The application processes the visual feed, proximity sensor's trigger, the gap between objects and the sensor, the UAV's location, and its orientation. This data is used to dynamically create an obstacle map, allowing users to visualize the UAV's environment in real time.
[00064] In an embodiment, there are a myriad of sensor types available, ranging from vision and imaging, weather, radiation, proximity, pressure, and many more. A specific type of sensor falling under the Proximity Sensors category is the ultrasonic detector. The device uses high-frequency sound waves to ascertain the distance to an object, converting the sound that returns into an electrical signal. Ultrasonic waves travel faster than audible sound waves. The ultrasonic sensor consists of two critical components: the source, which generates the sound via piezoelectric crystals, and the receiver. The sensor uses the transceiver to calculate the time elapsed between the generation of the sound and its contact with the receiver, which in turn is used to estimate the distance between the sensor and the object. The formula used is G = 12 T x S, where G represents the gap, T stands for time, and S is the speed of sound, which is 343 meters per second. By offering location data, communication can be facilitated between the sensors of the drone and an Android application hosted on the server.
[00065] This embodiment suggests potential future implementations of the method, where actual UAVs or multiple UAVs are deployed in a real environment. This approach would allow for observation of the UAVs' adaptive obstacle avoidance behavior in a realistic setting, potentially leading to further improvements and adaptations of the method.
[00066] Fig. 3 illustrates a network overview of adaptive neural control, in accordance with an embodiment of the present disclosure. As illsutrated, the autonomous behavior of the UAV system, also known as a drone, is achieved through the integration of an automaton, neural components, synaptic plasticity, and neurological feedback from the surroundings, creating a sensorimotor loop as depicted in Fig. 3. The neural dynamic loop communicates yaw and pitch information from the Robot Operating System (ROS) to the automaton which is reflected in the V-REP simulator. A flexible neural sensory processing system embedded with neural components and plasticity handles the information transfer and processing within the loop. Only the yaw and pitch information is relayed to the automaton for this study as it focuses solely on constructing a UAV system that operates at a specific altitude, hence pitch adjusts upward and downward movements while yaw steers left and right turns. The drone will rotate clockwise about its axis when the yaw value of the rate is positively evaluated, and conversely, it will move counter-clockwise when the value is below zero. For obstacle detection, two ultrasonic distance sensors are installed at the front of the UAV. Each sensor has a beam angle of ten degrees and a range of 50 centimeters, and they are positioned five centimeters away from the drone's forward motion axis at an angle of 35 degrees. The extent of elevation is kept constant as it's not the primary focus of the investigation. The key system methodologies encompass synaptic plasticity, the sensory system, the neural controller, and web services.
[00067] Fig. 4 illustrates a neural control arrangement system, in accordance with an embodiment of the present disclosure. Human bodies are constantly executing an array of functions simultaneously. Each neuron in our body fires approximately two hundred times per second. The brain, which serves as the human body's processing unit, manages the coordination. The neural network, a multi-layered network of neurons, has myriad applications. For instance, the neural network was employed to guide the drone and enable it to circumvent obstacles. A two-neuron repeating system with constrained uncertainty exhibits a range of intriguing dynamical properties, applicable to several tasks such as sensorimotor processing, scene recall, and control execution. A neural system, depicted in Fig. 2, is used to set up the control system of a drone for obstacle avoidance. All neurons within the system are characterized as discrete-time, non-spiking neurons. The computation for each 'ai' is displayed in Equation (1).

In Equation (1), 'Bi' signifies a major impact on neuron 'i', while 'Wij' denotes the synaptic quality of the relationship between neurons 'i' and 'j'. Except for the N6 and N7 neurons (as illustrated in Figure 2), which use a sigmoid function, the output of all neurons is calculated using a hyperbolic tangent (tanh) transfer function. Two separate ultrasonic sensors on the drone's front relay obstacle detection signals to both neurons. The raw sensor data, which ranges between -1 (indicating no obstacles within the sensor's range) and 1 (signifying a close obstacle), are processed and then relayed to the neural system. Nodes N1 and N2, with dynamic loads W11, W22, and W12, operate as asynchronous neurons with flexible transmitters. The output of these neurons, excluding N6 and N7, is calculated using the hyperbolic tangent function. The self-associations (W11 and W22) can generate a cyclic effect in their neural initiations, allowing the drone to maintain a continuous turning behavior, even in the absence of detected obstacles. The alternating output from neurons N1 and N2, where a positive yield from one results in a negative yield from the other, is significant for balance. These neurons form an even loop, which helps the drone avoid deadlock scenarios where both sensors simultaneously detect obstacles.
[00068] In an embodiment, synaptic plasticity, a vital neurological process, enables the brain to adapt and learn new information. It governs the changes in synapses - connections between neurons that facilitate communication. The efficiency of interaction between two neurons is dictated by synaptic plasticity, essentially controlling the 'volume' of their dialogue. The 'volume' isn't fixed, but instead, fluctuates over short and long time scales. Short-term synaptic plasticity refers to changes in synaptic strength happening within a second. For instance, swift alterations in this 'volume' help highlight crucial connections for communication. Conversely, long-term synaptic plasticity spans from minutes to hours, days, or even years. It's primarily responsible for content storage in the brain, aiding in the formation and interpretation of new mental images. The modification of each synaptic weight relies on three key elements: the system's output action Oi(t) at time 't', the previous output action Oi(t-1) at time 't-1', and a reflective signal Ri(t) provided in Eq. (2).

The response signal serves a crucial role in managing the operational setup of the research that commences when the drone identifies an obstacle within close range (approximately 30 cm). In addition to this, synaptic scaling can be employed to enhance the synaptic weights. For the learning model, the positive range vi [0, 1] is equated to the outputs Oi? [1, 1] to guarantee that the signs of the synaptic weights remain unaltered.

The self-association weights would be determined as Eq. (4) and Eq. (5):

Here, µr is the learning rate that assists in calibrating the timeline of the correlation learning process and is fine-tuned to 0.01. The forgetfulness rate, which modifies forgetting/synaptic scaling, is established at 0.0003 in the equation. To ensure a consistent decrease in weights when no obstacle is detected, k is reduced and set to -0.01. These modifications apply to the two avoidance connections (W12, 21) and are made identical to preserve symmetry. The procedure is concluded using temporary variables (q1, 2) which evaluate the average of the inhibitory synapses. The updates for the following variables are specified in Equations (6), (7), and (8):

In this context, µq refers to the learning rate and is set at 0.015. Meanwhile, ? and k represent the forgetfulness rates and the equilibrium respectively. All these meta parameters are empirically determined. As for controlling pitch, a sigmoid transfer function is employed to calculate the outputs of neurons N6 and N7, effectively mapping the output within the range of [0, 1]. This is crucial because, if an obstacle is close, there may be a need to decrease the pitch to stop the UAV, but negative outputs from the two neurons (N1 and N2) could prevent this. By assigning weights of 0.5 to the signals to N5, an average is achieved. Hence, if both sensors detect a nearby obstacle, the UAV is inclined to move in the opposite direction.
[00069] Fig. 5a and 5b illustrates a Client-server architecture of web-services and a RESTful communications between user and UAV/drone, respectively, in accordance with an embodiment of the present disclosure. Web services are popular tools utilized to enable interaction between devices of different architectures. They play a significant role in facilitating internet communication, often through a client-server setup. Web services primarily fall into two categories: RESTful web services and SOAP web services. SOAP web services communicate response messages exclusively in XML format. On the other hand, RESTful services utilize both XML and JSON formats. In the present disclosure, the system has opted for the JSON format due to its simplicity and speed. RESTful architecture emphasizes resources, with every component of the architecture regarded as a resource. These resources are accessed and manipulated via a common interface using HTTP standard methods. HTTP methods employed in REST-based architecture include: GET for read-only access to a resource, POST for creating new resources, DELETE for eliminating a resource, and PUT for updating or creating a resource.The system has implemented RESTful web services, an architectural style introduced in 2000 by Roy Fielding, for communication in our web service development. Owing to its efficient use of the internet, RESTful architecture is often dubbed as the language of the internet. The system has introduced a RESTful-based application programming interface to enable communication between the UAV or drone and the client application.
[00070] Fig. 6 illustrates simulation of obstacles map, in accordance with an embodiment of the present disclosure. As illsutrated, the simulation of an obstacle map refers to the digital representation of the environment in which a UAV or drone is operating. It displays potential obstacles within the drone's vicinity, allowing for predictive and reactive navigational adjustments. Such a map is typically constructed using data from sensors, such as lidar or ultrasonic sensors, that detect objects in the environment. The V-REP mapping simulation is crucial for ensuring safe and efficient navigation, particularly in complex or unfamiliar terrains. It enables the UAV to anticipate and avoid obstacles, thereby preventing potential crashes and improving operational efficiency.
[00071] Fig. 7 illustrates an Android applications showing visual feed from the drone, in accordance with an embodiment of the present disclosure. As illustrated, Android applications that show visual feeds from drones provide real-time video footage captured by the drone's onboard cameras. This allows users to remotely view the drone's perspective, essentially serving as its eyes in the sky. Such applications enhance situational awareness, aid in navigation, and enable a variety of uses, from aerial photography and videography to inspection tasks, surveillance, and exploration in areas difficult or dangerous for humans to reach.
[00072] Example embodiments herein have been described above with reference to block diagrams and flowchart illustrations of methods and apparatuses. It will be understood that each block of the block diagrams and flowchart illustrations, and combinations of blocks in the block diagrams and flowchart illustrations, respectively, can be implemented by various means including hardware, software, firmware, and a combination thereof. For example, in one embodiment, each block of the block diagrams and flowchart illustrations, and combinations of blocks in the block diagrams and flowchart illustrations can be implemented by computer program instructions. These computer program instructions may be loaded onto a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions which execute on the computer or other programmable data processing apparatus create means for implementing the functions specified in the flowchart block or blocks.
[00073] Throughout the present disclosure, the term ‘Artificial intelligence (AI)’ as used herein relates to any mechanism or computationally intelligent system that combines knowledge, techniques, and methodologies for controlling a bot or other element within a computing environment. Furthermore, the artificial intelligence (AI) is configured to apply knowledge and that can adapt it-self and learn to do better in changing environments. Additionally, employing any computationally intelligent technique, the artificial intelligence (AI) is operable to adapt to unknown or changing environment for better performance. The artificial intelligence (AI) includes fuzzy logic engines, decision-making engines, preset targeting accuracy levels, and/or programmatically intelligent software.
[00074] Throughout the present disclosure, the term ‘processing means’ or ‘microprocessor’ or ‘processor’ or ‘processors’ includes, but is not limited to, a general purpose processor (such as, for example, a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a microprocessor implementing other types of instruction sets, or a microprocessor implementing a combination of types of instruction sets) or a specialized processor (such as, for example, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), or a network processor).
[00075] The term “non-transitory storage device” or “storage” or “memory,” as used herein relates to a random access memory, read only memory and variants thereof, in which a computer can store data or software for any duration.
[00076] Operations in accordance with a variety of aspects of the disclosure is described above would not have to be performed in the precise order described. Rather, various steps can be handled in reverse order or simultaneously or not at all.
[00077] While several implementations have been described and illustrated herein, a variety of other means and/or structures for performing the function and/or obtaining the results and/or one or more of the advantages described herein may be utilized, and each of such variations and/or modifications is deemed to be within the scope of the implementations described herein. More generally, all parameters, dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and/or configurations will depend upon the specific application or applications for which the teachings is/are used. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific implementations described herein. It is, therefore, to be understood that the foregoing implementations are presented by way of example only and that, within the scope of the appended claims and equivalents thereto, implementations may be practiced otherwise than as specifically described and claimed. Implementations of the present disclosure are directed to each individual feature, system, article, material, kit, and/or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and/or methods, if such features, systems, articles, materials, kits, and/or methods are not mutually inconsistent, is included within the scope of the present disclosure.

Claims
I/We Claim:
1. An unmanned aerial vehicle (UAV) system for obstacle mapping and avoidance, comprising:
a UAV equipped with the proximity sensors, which generates the sensory data;
a two-neuron perpetual system network that processes the generated sensory data into the motor commands;
a correlation-based learning mechanism with a synaptic scaling for adjusting the synaptic weights within the network; and
an Android application for displaying the generated sensory data and dynamically creating an obstacle map.
2.The UAV system of claim 2, wherein the system is configured to adapt neuronal dynamics of the network based on different physical phenomena effects.
3.The UAV system of claim 2, wherein the system utilizes various angles formed by significant functions of the network to navigate intersection locations and dead ends.
4.The UAV system of claim 2, configured to mimic autonomous drone behaviour and obstacle avoidance strategies using a modified version of the V-REP Simulator.
5.The UAV system of claim 2, wherein the two-neuron perpetual system network applies post-processing neural units for pitch and yaw management to convert sensory information into motor commands.
6.The UAV system of claim 2, wherein the system's synaptic weights increase or decrease based on the perceived effectiveness of the UAV's perception of its surroundings, leading to adaptive neuronal dynamics.
7.The UAV system of claim 2, wherein the UAV is configured to inspect and navigate within a turbulent, unstructured environment by avoiding obstacles, sneaking around corners, and not becoming stuck.
8.The UAV system of claim 2, configured to adapt the synaptic weights within the two-neuron perpetual system network based on the UAV's perception of its surroundings, thereby enhancing the UAV's ability to recognize and avoid obstacles.
9.The UAV system of claim 2, wherein the varying angles formed by the network's significant functions enable the UAV to handle intersection locations and dead ends without becoming stuck.
10.A method for generating an obstacle map using an unmanned aerial vehicle (UAV), the method comprising:
deploying a UAV equipped with the proximity sensors to the detect physical objects and their proximity;
processing the sensory data through a synaptic plasticity-based repeating network, specifically a two-neuron perpetual system network, to derive motor commands;
adjusting the synaptic weights within the network based on a correlation-based learning mechanism and the synaptic scaling;
directing the UAV's pitch and yaw movements based on the motor commands to navigate and avoid the obstacles in the region; and
displaying the UAV's sensory data on an Android application to dynamically create an obstacle map.

UNMANNED AERIAL VEHICLE (UAV) SYSTEM FOR OBSTACLE MAPPING AND AVOIDANCE
Abstract
This invention discloses a unique system that utilizes an unmanned aerial vehicle (UAV) equipped with proximity sensors for autonomous obstacle mapping and avoidance. The UAV processes sensory data through a two-neuron perpetual system network, transforming it into motor commands. The network employs a correlation-based learning mechanism and synaptic scaling to adjust synaptic weights, thereby enhancing navigation and obstacle recognition capabilities. Moreover, the system integrates an Android application to display the UAV's sensory data, facilitating real-time mapping of the environment. The UAV system can adapt neuronal dynamics based on different physical phenomena, uses the network's functions for efficient navigation, and mimics autonomous drone behavior via a modified V-REP Simulator. The invention also discloses a method for generating obstacle maps using the UAV system, promising a robust solution for navigating turbulent, unstructured environments and providing dynamic obstacle mapping. , Claims:Claims
I/We Claim:
1. An unmanned aerial vehicle (UAV) system for obstacle mapping and avoidance, comprising:
a UAV equipped with the proximity sensors, which generates the sensory data;
a two-neuron perpetual system network that processes the generated sensory data into the motor commands;
a correlation-based learning mechanism with a synaptic scaling for adjusting the synaptic weights within the network; and
an Android application for displaying the generated sensory data and dynamically creating an obstacle map.
2.The UAV system of claim 2, wherein the system is configured to adapt neuronal dynamics of the network based on different physical phenomena effects.
3.The UAV system of claim 2, wherein the system utilizes various angles formed by significant functions of the network to navigate intersection locations and dead ends.
4.The UAV system of claim 2, configured to mimic autonomous drone behaviour and obstacle avoidance strategies using a modified version of the V-REP Simulator.
5.The UAV system of claim 2, wherein the two-neuron perpetual system network applies post-processing neural units for pitch and yaw management to convert sensory information into motor commands.
6.The UAV system of claim 2, wherein the system's synaptic weights increase or decrease based on the perceived effectiveness of the UAV's perception of its surroundings, leading to adaptive neuronal dynamics.
7.The UAV system of claim 2, wherein the UAV is configured to inspect and navigate within a turbulent, unstructured environment by avoiding obstacles, sneaking around corners, and not becoming stuck.
8.The UAV system of claim 2, configured to adapt the synaptic weights within the two-neuron perpetual system network based on the UAV's perception of its surroundings, thereby enhancing the UAV's ability to recognize and avoid obstacles.
9.The UAV system of claim 2, wherein the varying angles formed by the network's significant functions enable the UAV to handle intersection locations and dead ends without becoming stuck.
10.A method for generating an obstacle map using an unmanned aerial vehicle (UAV), the method comprising:
deploying a UAV equipped with the proximity sensors to the detect physical objects and their proximity;
processing the sensory data through a synaptic plasticity-based repeating network, specifically a two-neuron perpetual system network, to derive motor commands;
adjusting the synaptic weights within the network based on a correlation-based learning mechanism and the synaptic scaling;
directing the UAV's pitch and yaw movements based on the motor commands to navigate and avoid the obstacles in the region; and
displaying the UAV's sensory data on an Android application to dynamically create an obstacle map.