Method of assessing an expected aggregate accident severity for a plurality of persons during a standard time period

A computer-implemented method for computing and displaying an expected aggregate accident severity for a user over a time period, the method comprising the steps of:

• synchronizing, at a network-connected server, a plurality of sensor readings from a network-connected wrist wearable health-tracking device associated with a user, the network-connected wrist wearable health-tracking device comprising:
• an accelerometer and a three-axis gyroscope providing signals for determining wrist and arm motion to detect sleep;
• and
• a body temperature sensor, a pulse monitor, and a blood oxygen sensor providing signals to help determine stages of sleep;
• modeling, by the server, circadian oscillators for the user based on the plurality of sensor readings, wherein the circadian oscillators comprise a first oscillator having a period of 24 hours and a second oscillator having a period of 12 hours, the first and second oscillators being out of phase to model variations in arousal level wherein the modeling implements the circadian oscillators by summing two cosine waves, the first cosine wave completing one cycle every 24 hours with phase centered at approximately 1700 hours, and the second cosine wave completing two cycles every 24 hours with half the amplitude and offset by approximately three hours earlier, the sum producing a composite waveform that represents the user's circadian rhythm based on the time of day extracted from the sensor readings;
• determining, by the server, a sleep pattern of the user during a sleep period based on the plurality of sensor readings, wherein determining the sleep pattern comprises:
• detecting sleep state transitions by analyzing changes in wrist and arm motion from the accelerometer and gyroscope signals, wherein reduced motion below a threshold indicates sleep onset and increased motion above a threshold indicates wake events;
• tracking sleep quality by combining the motion data with physiological signals from the pulse monitor, temperature sensor, and blood oxygen sensor to identify sleep stages and disruptions;
• generating the sleep pattern as a time-series record of sleep states, sleep stages, and fragmentation events over the sleep period;
• calculating, by the server, an amount of effective sleep for the user based on the determined sleep pattern, wherein the effective sleep represents a cumulative measure of sleep recovery that increases during detected sleep periods and decreases during periods, the measure being adjusted based on sleep quality indicators from the sensor reading;
• generating, by the server, a fatigue assessment for the user during a work period subsequent to the sleep period by combining the determined sleep pattern with the calculated amount of effective sleep and the current circadian phase, wherein the fatigue assessment indicates performance capability based on the physiological recovery tracked through the sensor readings;
• generating, by the server, a real-time safety monitoring metric for the user based on correlation between the fatigue assessment and sleep pattern disruptions, wherein the metric quantifies deviations from baseline performance capability as measured through the continuous physiological monitoring;
• automatically adjusting, by the server, safety protocols in response to the real-time fatigue assessment, including generating alerts when sensor-detected fatigue patterns indicate reduced performance capability below a safety threshold;
• and
• displaying, on a graphical user interface of the network-connected user device, the safety monitoring metric with visual alerts that change based on threshold comparisons of the real-time fatigue assessment;
• repeating the synchronizing through displaying steps for each user in the plurality of users to generate aggregate safety monitoring metrics for the plurality of users during the time period based on the fatigue assessments of each user.