First Year: 2015 Cadre Research Topics
See the Participants Explain the Work in Their Own Words
Nicholas performed work on model-based software engineering (MBSE) techniques and applying them to the challenge of architecting and validating a CubeSat software system. The work centered on functional and quality requirements.
Zachary worked on a compression and decompression algorithm to enable the transfer of image data to/from orbit. This work, in conjunction with the aforementioned super-resolution work, focuses on ensuring that the compressed and subsequently decompressed data is still useful for various image processing applications.
Delia evaluated the impact of compression on the results obtained from super-resolution algorithms. The goal is to determine the trade-offs as compression is required to move significant amounts of image data from a small spacecraft in orbit to the Earth.
Connor performed work on the development of an intrusion detection system for use onboard a spacecraft. In particular, his work focused on enhancing an outlier detection algorithm to make it run at a suitably fast speed for this purpose.
Samuel worked on maintaining the confidentiality of collected data. A technique based on pixel shuffling was assessed.
Emily looked at how pulsed plasma thrusters and other similar propulsion systems could be used to de-orbit a spacecraft.
Michael worked on and evaluated technologies to enhance the integrity of transmitted data against natural impediments and to facilitate the detection of corrupted or altered data.
Alexander performed work on determining object range from a set of cameras that could be placed onboard a CubeSat or cluster of CubeSats. A correlation approach is being utilized and enhanced for this purpose.
Matthew worked to implement and enhance a design for an attitude determination and control system (ADCS) based on artificial (computational) intelligence principles. This system will learn and refine the spacecraft movement model based on a training regime and mission activities.
Andrew worked with neural networks for pattern recognition. He used convolutional neural networks to assess their capability to detect particular types of objects and animals in satellite imagery (using aerial UAV imagery for this work).
Wentong performed work aimed at determining whether satellite imagery was susceptible to adversarial attacks where small, barely human-perceptible changes could cause invalid identification or misidentification.