Ossi Pirinen and Davide Falanga

Autonomous systems are only as capable as their ability to reliably perceive and understand their surroundings. At the center of that capability is the imaging system: the cameras, sensors, image-processing pipelines, and perception algorithms that transform photons into actionable information. Drawing on the challenges of building vision-based autonomous aerial robots, this talk will explore the interplay between imaging hardware, real-time computer vision, and autonomy, and how co-designing these systems can improve robustness, expand the operating envelope, and enable new autonomous capabilities in complex, dynamic environments.

At Skydio, this relationship between imaging and autonomy is fundamental to how we build our products. Our autonomous drones rely on sophisticated camera systems not only to capture high-quality imagery, but also to continuously perceive, interpret, and navigate the world around them. By developing imaging, perception, and autonomy as an integrated system, Skydio translates advances in computer vision into real-world capabilities that enable aircraft to operate reliably in challenging environments and help customers perform critical tasks more safely and efficiently.