Cognitive workload classification of law enforcement officers using physiological responses

Area of Science:

• Traffic safety research
• Law enforcement operational efficiency
• Human-computer interaction in vehicles

Background:

• Motor vehicle crashes (MVCs) are a primary cause of fatalities among U.S. law enforcement officers (LEOs).
• Novice LEOs (nLEOs) experience significant cognitive workload while driving, increasing crash risk.
• Secondary tasks during patrol operations exacerbate cognitive load for LEOs.

Purpose of the Study:

• To develop a machine learning algorithm (MLA) for estimating LEO cognitive workload.
• To assess the feasibility of using physiological data for real-time workload monitoring.
• To inform the design of adaptive in-vehicle systems for enhanced officer safety.

Main Methods:

• A ride-along study involving 24 novice LEOs (nLEOs) was conducted.
• Physiological responses (heart rate, eye movement, galvanic skin response) were recorded using unobtrusive sensors.
• A random forest algorithm was employed to analyze physiological data and predict cognitive workload.

Main Results:

• The random forest MLA achieved over 70% accuracy in predicting cognitive workload.
• The algorithm's performance was based solely on physiological signal data.
• This demonstrates the potential of physiological monitoring for workload assessment.

Conclusions:

• A validated MLA can estimate LEO cognitive workload in real-time.
• This technology can enable adaptive in-vehicle systems to mitigate risks.
• Reducing cognitive workload can decrease the incidence of fatal MVCs in law enforcement.