Maximizing the battery life of a swarm of drones

Consider this not-too-far-fetched scenario: a person of interest involved in a law-enforcement hunt exits a building in a densely populated urban area and gains access to the public transportation system.

A swarm of drones hovers above the metro system, working in coordinated layers to track the subject without disrupting public safety. High-altitude drones scan the entire area, using advanced imaging and AI to map movement patterns and detect anomalies in the dense city landscape. At lower altitudes, agile drones focus on the target’s movements, weaving through urban corridors to maintain visual contact as he exits and re-enters different transit hubs. As he boards a train, the swarm anticipates possible routes, repositioning along key metro stops and deploying drones near exits to ensure no escape.

The lower-altitude drones seamlessly adjust, tightening their formation while the higher-altitude drones continue providing a wide-area overview. Working in concert, the swarm maintains relentless surveillance, adapting in real time to track every move with precision.

A swarm of drones can revolutionize modern-day problem-solving across various fields by leveraging autonomous coordination and real-time data processing.

In the military, drone swarms enhance reconnaissance, surveillance, and battlefield operations by providing rapid intelligence and overwhelming adversaries with synchronized attacks.

For security at large public events, they offer aerial monitoring, detecting threats, and ensuring crowd safety with minimal human intervention.

In traffic control, drones can dynamically assess congestion, optimize signal timings, and assist in accident response.

Additionally, drones play a crucial role in disaster relief by mapping affected areas, delivering emergency supplies, and locating survivors in hard-to-reach locations. Whether it is environmental monitoring, precision agriculture, or infrastructure inspections, drone swarms enable faster, more efficient, and cost-effective solutions to complex challenges.

Battery life is a major bottleneck for the utility of drone swarms, limiting their endurance and operational range. Most commercial and military drones rely on lithium-ion or lithium-polymer batteries, which provide only 20 to 60 minutes of flight time before requiring recharging or battery swaps.

An onboard video camera significantly impacts a drone's battery life. High-resolution cameras, particularly those capable of real-time streaming or thermal imaging, require substantial energy to process and transmit video data.

The more advanced the camera system—such as those with night vision, zoom capabilities, or AI-driven object recognition—the greater the drain on the power supply.

To mitigate this, drones often balance image quality with battery efficiency, using optimized compression algorithms, low-power processing units, or offloading computation to ground-based systems. However, in most cases, a drone equipped with a high-performance camera will experience shorter flight durations than one without.

The problem of balancing camera quality with battery efficiency in a swarm of drones is even more challenging, as not all the drones in a given swarm are necessarily equal. That is, some “drone’s pixels” may be more important than others, given the coverage it provides, the resolution of the pixels, their quality, and their utility (e.g., content) to the overall problem.

That is, a drone operating a 1 Mega-Pixel camera at 10 frames per second, hovering at low altitude with with a Field of View (FoV) obstructed by a tall building, may not provide pixels as valuable to the overall tracking problem as a drone operating a 10 Mega-Pixel camera at 20 frames per second, with a wider FoV.

Alternatively, a drone with a unique vantage point providing coverage that is not common to any of the drones in the swarm may be more important than a drone that is providing redundant coverage.

Therefore, a method that takes advantage of pixel-drone utility could adaptively optimize and orchestrate the image acquisition and transmission strategy so that the battery life of the swarm is maximized for this task of interest.

This is precisely what we address in our set of papers on maximizing the lifetime of imaging sensor networks (aka, swarm of drones)The challenge in this set of papers addressed is the large amount of data generated by multiple imaging sensors, which can quickly drain power and bandwidth resources.

There, we present a structured approach to optimizing how a network of drones—or imaging sensors—acquires, processes, compresses, and transmits images efficiently.

We propose a collaborative swarm strategy, where drones (or sensors) intelligently acquire, compress and share only essential data rather than transmitting redundant images. By exploiting both intra-sensor (within a single drone's view) and inter-sensor (overlapping views from multiple drones) correlations, the system reduces the total data load. The drones then utilize an optimal multi-hop routing method to transmit the compressed images in an energy-efficient manner, ensuring the network operates for the longest possible duration.

In a real-life drone swarm scenario, such a system could be highly valuable for persistent aerial surveillance, search-and-rescue missions, or military reconnaissance.

Instead of every drone continuously transmitting high-resolution video—which would rapidly deplete battery life—each drone could share only unique and critical image data, offloading redundant information.

A networked control system could then coordinate which drones should remain active, which should relay data, and which should conserve energy.

This approach would allow a swarm to track moving targets, monitor large events, or map disaster zones for extended periods without requiring frequent battery swaps or data overload, ultimately enhancing the effectiveness of drone-based operations.