Student working on a research project.

Advancing Safety Surveillance Using Individualized Sensor Technology

Work-related injuries represent a significant economic burden and involve substantial adverse personal outcomes. Worker fatigue has been identified as a risk factor for both acute and cumulative injuries. Significantly reducing the incidence of fatigue-induced workplace injuries depends on the accurate and timely detection of fatigue portents. The main objective of the proposed work is to develop a sensor-based, real-time exposure assessment system to determine deviations in measured physical and physiological indicators of fatigue. To achieve this objective, three specific aims are proposed. First, the appropriate combination of sensors and on-body locations for real-time fatigue monitoring will be identified using an experimental study. Specifically, participants (from industry) will complete a series of simulated manufacturing tasks while instrumented with three prototype sensor systems to capture their physical (accelerations, postures) and physiological (heart rate, perspiration) responses. The accuracy and reliability of the system will be compared to gold standard measurements for selection of the optimal sensor and location combination. Second, fatigue development models that distinguish between normal and fatigued states will be constructed. This aim will encompass the development of a) sensor fusion techniques for an aggregate estimate of the fatigued state of a worker; b) change point analysis (control charts) to define a threshold for fatigue identification; and c) diagnostics to determine the root cause(s) of fatigue. Third, a framework for classifying interventions for a given fatigue development profile will be developed and used to identify optimal interventions. A database of interventions will be populated based on stakeholder input. Taxonomy for these interventions will be created based on a Pandora-like classification system that combines expert knowledge and the attributes from the fatigue development model. The appropriateness of the intervention selection approach will be evaluated through an experimental study. The integration of knowledge from safety science, sensing technology, industrial statistics, and machine learning will create a platform for individualized, evidence-based exposure assessment. This will enable safety practitioners to intervene at the onset of fatigue, thus minimizing the potential for adverse personal outcomes and safety incidents. This work directly contributes to several strategic goals within the National Occupational Research Agenda.

Research Project Information

Disciplines: Industrial Engineering, Ergonomics, Biomedical Engineering, Exercise Science
Student Skill-Set Needed: data collection and analysis, programming
Compensation: Academic Credit, Volunteer
Available: Fall, Spring, Summer


For further information on this opportunity, or to apply, contact:

Faculty Member: Lora Cavuoto
Title: Assistant Professor
Department: Industrial And Systems Engineering
Office: 324 Bell Hall
Phone: 7166454696