Challenges of Autonomous Drones

Domain-specific challenges in drone autonomy include energy constraints like limited battery life, computational limitations on board for processing, environmental factors such as adverse weather, and operational complexity in dynamic environments like urban areas. Other challenges are the need for robust sensors and software, especially in GPS-denied environments, and navigating through dynamic obstacles and unpredictable conditions. All these challenges can be divided into the following three groups.

• Energy constraints: Battery life limits flight duration and mission scope, which is a significant challenge for long-distance flights and complex tasks.

• Adverse weather: Precipitation, fog, and strong winds negatively impact a drone's visibility, endurance, sensors, and ability to navigate and avoid collisions.

• Payload and weight restrictions: Weight limits the size and power of onboard computers and other critical components, forcing a trade-off between computational power, payload capacity, and flight time.

• Onboard computational power: Tasks like real-time image processing, localization, mapping, and decision-making require significant processing power, which is difficult to fit on a small drone due to weight and power constraints.

• Dynamic environments: Navigating and safely avoiding obstacles in environments with moving objects, such as crowds or other vehicles, poses a constant challenge in route planning.

• Software and security: Developing and testing software that is functionally safe, secure, and can handle unexpected situations on its own is a major hurdle.

• Sensor and perception limitations: Drones need reliable sensors (like cameras and radar) that can work across a wide range of conditions, but these sensors can be susceptible to weather influences, and their performance can be inconsistent.

• Regulatory hurdles: The creation of new rules, air traffic control systems, and predefined routes is necessary but lags behind technological advancements.

• GPS-denied navigation: In environments where GPS is unreliable or unavailable (e.g., indoors, under dense vegetation), drones must rely on other methods for navigation.