The Bronco Ember technology is designed to autonomously detect, track, and log nascent wildfires or similar phenomena. Bronco Ember consists of a custom two degrees of freedom (2DOF) precision pointing system and edge computing system for onboard artificial intelligence / machine-learning acceleration. This flight demonstration will be observed in the short-wave-infrared (SWIR) wavelength range.

We will demonstrate the capabilities of classical, quantum and hybrid machine learning onboard a CubeSat. With our array of sensors, we are able to evaluate quantum/classical machine learning approaches on Earth observation tasks, including atmospheric gravity wave measurements or multispectral image classification and segmentation. We will demonstrate how on-board data processing and machine learning result in reduced downlink bandwidth.

Using visible and infrared cameras, our payload automatically identifies and classifies plumes in Earth’s atmosphere using convolutional neural networks (CNNs). It then actively tracks identified plumes and saves images for operator analysis.