Principal Investigators: Dr. Geoffrey Landis and Anthony Colozza
Research Associates: Nathan Boll, Denise Salazar, and Christopher Stelter
Venus is a scientifically fascinating planet, and one that, due to its similarity and differences from Earth, is an interesting target of exploration, but to date Venus is far less explored than our neighbor planet Mars. Yet it would be an exciting goal, since Venus is an unknown planet, a planet with significant scientific mysteries, and a planet larger than Mars with equally interesting (although less well known) geology and geophysics. A mission entering and exploring the atmosphere of Venus would expand our knowledge of the environment of terrestrial planets.
This project will look at conceptual design of a mission to Venus. In particular, we will be interested in methods of exploring the upper atmosphere of Venus using a remotely-piloted aircraft. If a potential mission concept is established as feasible the identified concept will undergo further development and refinement of details. Otherwise, alternative mission concepts will be pursued and analyzed. Students will work with a mentor to research design concepts, setup up the science objective of the mission, and analyze the design. A report detailing the characteristics and feasibility of the concept will be compiled and published.
Secure Autonomous Data Mule Project
Principal Investigators: Dr. Dennis Iannicca and Alan Hylton
Research Associates: Anthony Bentley and David Riesland
Historically, lunar and planetary exploration spacecraft have been few in number and have communicated only with operations centers on Earth. Such communications have in effect been dedicated interplanetary communication circuits, established and configured by human operators. Recent experiences at Mars have shown that using orbiters to relay data from the surface can greatly increase science data return; similar benefits are expected in other locales such as around the moon. International cross-support for this relaying capability will increase mission robustness and science data return, allowing surface elements multiple opportunities to forward data to Earth. As the number of relays and relay users increases, the need to automate a standard data relay service will increase.
The familiar Internet network protocol model is not suitable for this task, as it is not designed for effective operations over communication links characterized by very long signal propagation latencies, frequent and prolonged service interruptions, and limited and highly asymmetrical transmission rates.
Communication with and among a large and growing population of communicating entities (robotic sensors, for example) separated from Earth by interplanetary distances and/or by recurring lapses in mutual visibility due to orbital or planetary motion will require deployment of a store-and-forward communication network that is capable of providing reliable data delivery and dynamic routing in a fully automated fashion.
The objective of the project is to develop and demonstrate a secure autonomous data mule system utilizing Disruption Tolerant Networking (DTN), dynamic routing, and opportunistic contacts. The system will be comprised of a group of sensor nodes, a base station, and one or more autonomous robotic data mules that will move from node to node collecting the sensor data for delivery to the base station. All communications in the system must be authenticated and securely communicated from the sensor nodes to the base station. The data mule should be designed to avoid simple obstructions in its path and be able to calculate the most efficient alternate path to collect the data in the event of an obstructed path in order to demonstrate DTN’s opportunistic contact capability rather than relying on scheduled contacts. Once the data is uploaded to the base station it will be transmitted over a simulated multi-hop Deep Space Network back to Earth. A team of 2-3 students will work with mentors to research appropriate configurations and designs, setup equipment, and demonstrate the system in a relevant way.
Real-time, multi-modality, multiple cognitive state monitoring to improve aviation safety
Principal Investigators: Dr. Beth Lewandowski, Dr. Tristan Hearn, and Angela Harrivel
Research Associates: Nikhil Garg, Elizabeth Pickering, and Kier Fortier
The cognitive state of a person is constantly fluctuating with time and situation. Examples of different possible cognitive states include fatigued, bored, distracted, engaged, engrossed, confused, anxious or stressed. Ideally everyone would always be alert, engaged and right on task while performing required duties. However, the practical situation is often less than ideal and could include someone struggling to perform duties because they are sleep deprived or distracted. These less than ideal cognitive states can lead to operational accidents. NASA has a goal of improving aeronautical safety by applying cognitive state monitoring during piloting duties as part of the Vehicle Systems Safety Technologies Project, which is part of NASA’s Aviation Safety Program. This work is relevant to the mitigation of risk due to human performance decrement in any safety-critical task.
The objective of the project is to use multiple methods to measure physiological signals, such as heart rate, electroencephalography and blood oxygenation with functional near infrared spectroscopy, and to develop classification schemes based upon the signals that can be used to detect cognitive state. Behavioral data is also collected, and serves to validate the induced state. The long term goal is to perform the state monitoring in realtime. Therefore, methodology to perform signal processing in real-time is also needed. This multi-disciplinary work is very appropriate for a team project. The typical steps necessary for performing cognitive state classification using physiological signals include: 1) Development of an experimental task and protocol which puts the subject into a specific cognitive state; 2) Data collection while the experimental protocol is performed; 3) Signal processing including filtering, artifact decontamination and feature extraction; 4) Determination of a classification scheme and individual training; and 5) Prediction accuracy determination based on the true state as indicated by the behavioral data during a subsequent test or operational setting.