Advancement of edge computing and artificial intelligence (AI) in recent years has significantly altered how we use drones. The ability to decipher information locally, as compared to relying on distant servers makes drones increasingly smart and powerful due to these technologies. This implies that drones can operate in real time, regardless of the most isolated locations.
Collectively, we will examine the development of drone technology, the frontiers of AI and edge computing, and the implications for many use cases.
Pradeep Kirnapure, Distinguished Engineer at Hughes Systique Corp. (HSC), points out:
“Autonomous systems have evolved over the years and have significantly moved up the value chain with the help of AI technology enablers and edge computing. Some of the common examples of autonomous systems are self-driven cars, warehouse robots, unmanned aerial vehicles (UAV), unmanned underwater vehicles (UWV), etc. Given the popularity of the UAV (commonly referred to as drones) in both military and civil applications, this article considers drones as a specific example to highlight how AI and edge computing are contributing to the evolution of autonomous systems,”
Drone technology history
The early 1900s saw the introduction of basic remote-controlled aircraft, which marked the birth of drone technology. However, drones did not become advanced until the start of the Cold War. The US, for example, made significant advances with the Ryan Model 147 in the 1960s. According to Menkhoff et al. (2022), the major aim of early drones was surveillance, but they lacked the advanced electronics seen in modern drones.
UAVs using edge computing and AI
Drones grew increasingly advanced over time. Towards the end of the 20th century, they could carry more gear, have longer flight durations, and improve navigation systems. The framework for incorporating AI, which would transform drone capabilities, was established around this time. Drones were primarily military instruments at first, reliant on manual control systems and having little autonomy (Dai et al., 2022).
Initial applications out of military
Drone technology began in the early 1900s with basic remote-controlled aircraft. However, until the Cold War, drone technology remained very simple. For instance, the United States achieved significant progress in the 1960s with the Ryan Model 147. Menkhoff et al. (2022) claim that early drones were mostly used for surveillance and lacked the advanced technology found in modern drones.
AI and edge computing changing drones
Varinder Singh Jawanda, CEO at Millennium Automation Pvt Ltd (MAPL World), explains: “AI is about creating machines that can think and learn like humans. Edge computing means processing data close to where it’s generated rather than sending it to a far-off server. When you combine AI with edge computing, you get Edge AI, which allows drones to make smart decisions in real-time right from where they are.”
Processing data in real time
One of the biggest changes in drone technology is the ability to process data instantly. With edge AI, drones don’t need to wait for data to be sent to a server. Instead, they can analyze data right on the spot. For example, in emergencies like natural disasters, drones with edge AI can quickly assess damage and send important information to rescue teams without needing remote control.
A 2024 study published in IEEE Access highlights that drones equipped with edge computing can process data up to 10 times faster than those relying on cloud-based systems. This speed boost significantly reduces latency, which is crucial in scenarios requiring immediate data analysis, such as disaster response.
Better situational awareness
AI-powered drones with edge computing can gather and analyse data from sensors like cameras and LiDAR. This helps them understand their surroundings better, even in tough or remote conditions. They can create detailed maps, spot important objects, and notice unusual things, which is incredibly useful for many tasks.
Operating without remote control
Today’s drones can operate on their own, thanks to AI and edge computing. They can navigate, avoid obstacles, and gather data without constant human control. This is especially helpful in places where connectivity is poor or where manual operation isn’t practical (OpenMind, 2024, AI and Drones).
Pillars of real-time autonomous system performance
Pradeep Kirnapure, HSC, says: “Autonomous systems, such as drones, must perceive, analyze, and respond in real-time to operate effectively in the physical world. AI and edge computing is pivotal in enabling these functions,”
Perceiving: Drones use various sensors, such as cameras, radar, and Lidar, to generate large volumes of data, creating a 3D view of their surroundings. This data must be processed instantly using complex machine learning (ML) models onboard the drones, as relying on cloud servers would introduce unacceptable delays.
Analyzing: Once the task is understood, the drone analyses the environment and plans the optimal path, considering constraints like battery life and avoiding obstacles. Computer vision algorithms and ML models onboard help with object detection, recognition, and trajectory estimation. AI-based SLAM techniques are employed for localisation when GPS is unavailable.
Responding: The final step is executing planned actions with minimal human intervention. Drones use advanced AI/ML algorithms and onboard computing for tasks like flight control and collision avoidance. Relying on edge computing rather than cloud servers ensures rapid response times, crucial for tasks requiring sub-second decisions, such as avoiding collisions.
Varinder Singh Jawanda, MAPL World, states, “Looking ahead, as autonomous systems like AI-driven drones continue to advance, they bring both exciting opportunities and significant challenges.”
Technical and ethical issues
Despite their benefits, AI-driven drones face challenges. These include technical limitations, privacy concerns, and ethical issues. Ensuring these drones work reliably and securely while protecting data privacy is essential.
Policies and regulations
Regulations must advance with drone technology. Legislators must strike a balance between promoting innovation and attending to privacy and safety issues. Proper regulations will guarantee that drones are operated sensibly.
Pradeep Kirnapure, HSC concluded: “Indeed, AI/ML and edge computing have been significant contributors to advances in autonomous systems, considering the level of maturity, stability, and reliability we see today. Autonomous AI is carved as a specific area within the AI discipline to have a dedicated focus on the autonomous systems.”
In conclusion, AI and edge computing have dramatically improved autonomous drones, allowing them to process information in real time and operate independently. From their humble beginnings to their current high-tech uses, drones have come a long way. As technology advances, AI-driven drones will become even more integral in fields ranging from defence to environmental monitoring.
— Pradeep Kirnapure, Distinguished Engineer at Hughes Systique Corp. (HSC), and Varinder Singh Jawanda, CEO at Millennium Automation Pvt Ltd.
