TECHNOLOGY AREAS: Human Systems

ACQUISITION PROGRAM: Capable Manpower FNC (CMP-FY11-02)

OBJECTIVE: Develop lightweight, unobtrusive, modular, and wearable recording device(s) to capture, synchronize, and download environmental, physiological, physical, and subjective measures that contribute, and are associated with physical and cognitive fatigue. The device(s) or system(s) should also be capable of objectively and reliably assessing fatigue uncontaminated by individual factors such as aptitude, learning effects, yet sensitive to phenotypic differences in vulnerability to fatigue.

DESCRIPTION: Current Naval platforms and systems require increased operational capability with reduced manpower. The ability to accurately predict warfighter performance (e.g. accuracy and reaction time for job duty tasks) on these platforms and systems is an essential component of conducting cost, schedule, and performance tradeoffs between hardware, software, and human capabilities and limitations. These tradeoffs are increasing done with human performance models such as Total Crew Model and IMPRINT. While the capability to model warfighter performance has made great strides over the last several decades, models still lack the fidelity to support fine grained tradeoff analyses. Furthermore, most models are not validated, and have little capability to account for the impact of environmental stressors. The need to account for environmental stressors such as fatigue, motion, vibration and extreme temperatures is critical because they can result in physical and cognitive fatigue, leading to degradation in warfighter performance.

A significant challenge in validating human performance models is the ability to collect environmental and performance data from warfighters in an operational setting. Current methods are primarily paper based, although, standalone recording devices such as actigraph may also be used. Several limitations are associated with these current methods. First, from a participants� perspective, generating responses while performing mission tasks is cumbersome and time consuming. Second, from an experimenters� perspective, coding self-reported responses is time consuming and increases the likelihood of errors in data entry. In addition, experimenters may not be able to attend an experimental event in person (e.g. live fire testing) or collect all the environment conditions that a warfighter experiences (e.g. motion, vibration, noise). Third, data analysis cannot be performed until all data sheets are collected and coded thus delaying when the analysis is performed, and eliminating the possibility of real-time (or near real-time) analysis. Lastly, the lack of synchronization between devices and subjective reporting makes associations between objective and subjective data more difficult.

The end result of these limitations is that researchers focus more energy and time on collecting and processing a limited amount of data, then on assessing the impact of environmental stressors on fatigue. Accurately accounting for the effects of environmental stressors on operator performance will allow human performance models to better support assessment, analysis, and mitigation of stressors during system development, testing, and acquisition. To overcome the data collection challenges associated with validating human performance models, and ensure an accurate account for the effects of environmental stressors on performance, a novel, integrated, non-obtrusive data collection and analysis system is needed.

The selection of sensors needed for the data collection suite should capable of objectively and reliably assessing fatigue and based on theoretical models (Mallis, Mejdal, Nguyen, and Dinges, 2004). Recent advances in individualization algorithms are making possible a new generation of systems to tailor mathematical model-based assessments of fatigue to track trait-like differences among individual operators. (Van Dongen et al., 2007). Measures of fatigue that do not involve obtrusive or invasive physiological monitoring, that have high face validity for operator performance relative to vigilance based tasks, and that are free of contaminating factors such as aptitude and learning effects should be explored. These measures would be combined with state-of-the-art, mathematical model-based analysis and individualization algorithms to account for individual state and trait related performance changes due to fatigue (Kan et al, 2009; Mollicone, Van Dongen, Rogers & Dinges, 2008; Mott et al., 2009). The system would provide visualization tools based on behavioral alertness data to assess operationally relevant performance features sensitive to fatigue that are associated with phenotypic differences across individuals.

Academia and industry are increasingly developing portable and wearable data collection technologies. These systems are part of a larger effort of ubiquitous and pervasive computing applications and often have a set of typical sensors with them (Beigl, Krohn, Zimmer, and Decker, 2004). Some devices such as the iPod touch also have the capability of obtaining subjective responses via an embedded rating system. However, none of these devices nor technologies provide the capability for non-obtrusively collecting and then integrating data spanning environmental, physiological, physical, and subjective measures of fatigue. In addition, to these sensors it is necessary for designers to consider the usability of these systems. Based on prior work, Tapia, Intille, Lopez, and Larson (2006) developed four usability goals (i.e., ease of installation, ease of use, adequate longitudinal performance in natural setting, affordable for researchers) for a portable sensor kit that could be used for non-laboratory studies. Consideration of such usability guidelines is critical for the successful adoption of data collection tools.

PHASE I: Develop the framework for a data collection suite for measuring human performance. The components in the framework should build on and extend the state-of-the-art capabilities in portable and wearable sensors / devices and performance measurement (including environmental, physical, physiological, and subjective indices). All sensors / devices must be capable of collecting multiple sources of data simultaneously and synchronizing that data. The selection of sensors needed for the data collection suite should capable of objectively and reliably assessing fatigue and based on theoretical models (Mallis, Mejdal, Nguyen, and Dinges, 2004). In addition, the user software must allow data to be quickly downloaded onto a Windows based-PC, and reset the sensor / device for another data set. The components in the framework should work within an open systems architecture.

PHASE II: Develop a prototype suite of data collection tools based on the framework established in Phase I. Submit appropriate and necessary regulatory documents for testing using human participants. Validate the tools through empirical evaluations with the targeted user community.

PHASE III: Produce and market the suite of data collection tools for integration with ship and submarine test and evaluation programs.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: The suite of tools will have widespread applications to military, government, and private sector organizations in which it is important to assess performance (e.g., when fewer personnel are required to perform the same tasks and missions without degraded performance).

REFERENCES:
1. M. Beigl, A. Krohn, T. Zimmer, and C. Decker, "Typical Sensors Needed in Ubiquitous and Pervasive Computing," in Proceedings of the First International Workshop on Networked Sensing Systems (INSS '04). Tokyo, Japan, 2004, pp. 153-158.

2. M. Tapia, S. Intille., L. Lopez., and K. Larson., "The Design of a Portable Kit of Wireless Sensors for Naturalistic Data Collection", in Proc. Int'l Conf. Pervasive Comp. 2006: p. 117-134.

3. Van Dongen, HPA, Mott, CG, Huang, J, Mollicone, DJ, McKenzie, FD, and Dinges, DF, Optimization of biomathematical model predictions for cognitive performance impairment in individuals: Accounting for unknown traits and uncertain states in homeostatic and circadian processes. Sleep, 2007. 30(9): p. 1125-1139.

4. Kan, K, Mott, C, Van Dongen, H, Huang, J, Mollicone, D, and Dinges, D., Individualizing predictions of performance impairment across sleep/wake state transitions. in Aerospace Medical Association�s 80th Annual Scientific Meeting. 2009. Los Angeles.

5. Mallis, M.M., Mejdal, S., Nguyen, T.T., Dinges, D.F., Summary of the key features of seven biomathematical models of human fatigue and performance. Aviation, Space & Environmental Medicine 75(3):A4-A14, 2004.

6. Mollicone, DJ, Van Dongen, HPA, Rogers, NL, and Dinges, DF, Response surface mapping of neurobehavioral performance: Testing the feasibility of split sleep schedules for space operations. Acta Astronautica, 2008. 63(7-10): p. 833-840.

7. Mott, C, Van Dongen, H, Kan, K, Dinges, D, Huang, J, and Mollicone, D., Optimizing individual performance predictions in real-time scenarios using Bayesian estimation. in Aerospace Medical Association�s 80th Annual Scientific Meeting. 2009. Los Angeles.

KEYWORDS: portable and wearable systems; measurement; assessment; modeling; pervasive computing

Questions may also be submitted through DoD SBIR/STTR SITIS website.