Samsung Collaborates With the University of Georgia To Define and Measure Energy for Innovative Digital Health Experiences

Set for release in the second half of this year, the latest Galaxy Watch will incorporate One UI 6 Watch to further enhance the AI-driven health functionalities of the wearable. A standout feature of the new interface is the Energy Score,¹ which provides users with personalized health recommendations based on a numerical score reflecting their overall energy levels.

To develop the Energy Score feature, Samsung Research² collaborated with Professor Patrick O’Connor from the Mary Frances Early College of Education’s Department of Kinesiology at the University of Georgia, USA.

Samsung Newsroom explored the new Energy Score — designed to enhance the digital healthcare experience.

Energy is closely related to everyday efficiency and fatigue, and based on imperfectly understood neural circuit activity in the brain. As a result, objectively quantifying energy is not easy. Samsung Research and Professor O’Connor’s team joined forces to define and calculate an individual’s energy based on the available science.

A leading expert in exercise psychology, Professor O’Connor has extensively studied the effects of sleep, physical activity and caffeine intake on cognition and feelings of energy. He advised on the development of the Energy Score feature on Samsung Health, combining physical activity with neuroscience data and theory to suggest the concept of “Overall Capacity.”

While most existing services focus solely on physical aspects when measuring energy, Overall Capacity considers both physical and cognitive factors. Mental stressors can affect energy levels, as evidenced by disrupted sleep patterns and increased fatigue when stressed.

Essentially, energy reflects the amount of activity one can sustain relative to their total capacity. Exceeding one’s usual physical or mental load reduces energy in the short term. For example, if someone typically exercises at a low intensity for 30 minutes daily — but decides to engage in moderate-intensity exercise for an hour one day — their Energy Score is expected to drop the following day.

Regular exercise can improve Overall Capacity, potentially resulting in a higher Energy Score for the same workout intensity over time.

▲ Physical and cognitive loads that exceed average levels lead to reduced energy. However, regular exercise can increase energy or help maintain it at a higher average level.

Energy Score is calculated based on physical activity levels, sleep quality, heart rate during sleep and heart rate variability during sleep as measured by wearable devices. Samsung Research developed the Energy Score feature based on scientific studies and clinical research that studied the correlation between these indicators and cognitive, self-reported and physical markers of energy.

“Activity” can influence physical capacity. One can predict their energy for the day by comparing workout data from the previous day against their typical activity levels. This method employs the Acute:Chronic Workload Ratio concept that anticipates fatigue by assessing both long-term and short-term workload.

“Sleep” is primarily associated with mental capacity. Sleep patterns like average sleep duration over seven days, consistency of sleep and wake times and how quickly one falls asleep are analyzed and integrated into one’s Energy Score. This method follows the “energy reservoir” model — which explores the connections between sleep, fatigue and cognitive function, to calculate fluctuations in energy based on sleep duration and circadian rhythms.

“Sleeping Heart Rate” and “Sleeping Heart Rate Variability” are linked to both physical and mental capacities. Energy is forecast by comparing recent measurements to past long-term data trends. Predictions are more accurate when heart rate during sleep and heart rate variability during sleep data are both analyzed during a stable state of sleep.

▲ Indicators used to calculate the Energy Score. Activity and sleep are primarily associated with physical and mental capacity, respectively, while sleeping heart rate and sleeping heart rate variability are known to be associated with both.

Samsung Research took age and gender into account in determining how much weight each factor should have on one’s Energy Score. Furthermore, Professor O’Connor’s research team conducted experiments involving cognitive tests and self-reports of energy and fatigue symptoms, finding a significant correlation between the Energy Score generated by Samsung’s models and clinical data collected by the University of Georgia’s researchers.

Samsung Health’s Energy Score goes beyond simple numerical values. The figure offers health guidance and suggestions based on seven key factors affecting the score including average sleep duration and physical activity from the previous day.

To achieve this, Samsung Research combined its energy score model with optimization AI and generative AI technologies. Optimization AI first pinpoints key influences on the score and analyzes current energy levels along with recent lifestyle changes to suggest potential improvements. On-device generative AI then crafts these insights into friendly messages while upholding user privacy.

The health guidance provided gives users an understanding of their current Energy Score and science-based suggestions to manage appropriate levels of activity and rest for the day. By paying more attention to the factors that affect their daily Energy Score, users can go on to improve their lifestyle habits.

▲ Energy Score provides users with personalized health guidance via friendly, AI-generated messages

Professor O’Connor conducted extensive research in collaboration with Samsung developers, striving to enhance the reliability and validity of the Energy Score.

“From a scientific perspective, the Energy Score reflects predicted variation in the ability to perform brief cognitive tests of attention across the day based on objective information obtained from smart device sensors during the past week,” he explained.

“Through our collaboration with Professor O’Connor, we were able to address this challenge in a scientifically meaningful way,” said Yunsu Lee, Head of the Data Intelligence Team at Samsung Research. “We will continue to develop data and AI technologies to ensure that Samsung’s various devices are used more widely to enhance users’ lives.”

¹ Only for general health management and fitness purposes. Service availability may vary by country.
² The advanced research and development organization in Samsung Electronics’ DX (Device eXperience) division