| Affiliations: | College of Sciences |
| Team Leader: |
|
| Faculty Mentor: |
Mustapha Mouloua, PhD
|
|
Team Size:
|
5 |
| Open Spots: | 2 |
|
Team Member Qualifications:
|
Preferred Qualifications: Familiarity with reading and interpreting peer-reviewed research papers, particularly in areas such as human factors, physiological signal processing, python and simulators. Required Qualifications: CITI Training Human Subjects Research – Group 2: Social/Behavioral Research Investigators and Key Personnel Research and HIPAA Privacy Protections University Research Authorization Completion and approval of the URA (Undergraduate Research Authorization) form prior to participation. Time Commitment Ability to commit to in-person lab hours as required by the project. Reliable availability throughout the agreed-upon research period. Task Completion & Communication Commitment to completing assigned tasks by set deadlines. Openness to consistent communication with lab leadership and team members (email, shared platforms, meetings). Professional & Ethical Conduct Adherence to ethical standards for human subjects research. Respect for participant confidentiality and data security protocols. Organizational & Collaborative Skills Ability to follow protocols accurately and maintain organized records. Willingness to work collaboratively within a research team and accept feedback. |
|
Description:
|
This project proposes the development of a closed-loop adaptive AI system that detects early physiological and behavioral indicators of stress in space mission analog environments and dynamically reallocates tasks to autonomous systems before performance degradation occurs. Although prior research has demonstrated the ability to measure stress, model cognitive workload, and design adaptive automation systems, there is currently no validated framework that integrates early physiological stress biomarkers with predictive cognitive reliability modeling to proactively guide task redistribution in long-duration isolation settings. Using open-source data from space analog studies (e.g., NASA HERA, HI-SEAS, Antarctic overwintering research), machine learning models will be trained to identify patterns in heart rate variability, sleep disruption, reaction time drift, and task performance metrics that signal early cognitive strain. The system will generate a real-time stress index and simulate adaptive autonomy strategies, such as reducing multitasking demands or transferring non-critical tasks to automation when decreased reliability is predicted. By bridging stress detection and task-aware autonomy, this project aims to enhance human-autonomy teaming, resilience, and mission safety for future deep-space exploration. |