Hsin-Chieh Wu - Academia.edu (original) (raw)
Papers by Hsin-Chieh Wu
Ergonomics, 2011
This study explores the preferred viewing distance and character size for an electronic paper dis... more This study explores the preferred viewing distance and character size for an electronic paper display for three age groups. Proofreading speed and accuracy ratio were measured during Chinese proofreading tests using the preferred character size and minimum acceptable character size. Data analysis showed that the mean preferred viewing distance for young, middle-aged and older groups was 503, 455 and 444 mm, respectively. The mean preferred character size determined by young, middle-aged and older groups was 42.0, 50.0 and 55.2 min arc, respectively. The proofreading test results indicated that the older group proofread significantly more slowly (1.25word/sec) than the young (1.76 word/sec) and middle-aged groups (1.74 word/sec). Further, the participants proofread more correctly with their preferred character size (73.3%) than with their minimum acceptable character size (65.4%). This study provides valuable information for the design of Chinese text presentations for various agegroups. Statement of Relevance: This study confirmed the preferred viewing distance and character size for E-paper display were influenced by age. The preferred Chinese character size for young, middle-aged and older people was 42, 50 and 55 min arc, respectively. Therefore, the age factor should be considered for E-paper displays design and video display terminal (VDT) guidelines.
International Journal of Industrial Ergonomics, 2004
The semiconductor industry is moving from the production of 200 mm wafers to 300 mm wafers. With ... more The semiconductor industry is moving from the production of 200 mm wafers to 300 mm wafers. With the increase in wafer size, the workload of wafer handling tasks is also increasing. This study aims to evaluate the operator's handling capability and workload of handling 300 mm wafers by psychophysical, physiological and biomechanical approaches.
IEEE Transactions on Semiconductor Manufacturing, 2003
Semiconductor manufacturing industry is moving into the production of 300-mm wafers. To solve the... more Semiconductor manufacturing industry is moving into the production of 300-mm wafers. To solve the increased workload problem in manual wafer handling, some personal guided vehicles (PGVs) have been developed to help in the transfer of Front Opening Unified Pods (FOUP). This study compares two kinds of PGVs with a traditional cart and evaluates the feasibility of using them for manual FOUP handling tasks. Manual FOUP handling capability was assessed. The results indicate that there is no obvious advantage in using any of the two evaluated PGVs over the manual cart. There is potential risk of causing musculoskeletal disorders for female operators to handle the 300 mm FOUP manually. Since the development of a fully automated intra-bay FOUP handling system is a project of high technical difficulty, a combination of manual and automated handling is the current approach. To enhance the operator's health, safety and productivity, selection and training of operators, adequate design of handling tools and machine interface, assessment and balancing of workload are necessary.
Ergonomics, 2005
This study was aimed to investigate complete recovery time (CRT) after exhaustion in high-intensi... more This study was aimed to investigate complete recovery time (CRT) after exhaustion in high-intensity work. Twenty-four subjects were divided into two groups based on the cardiorespiratory capability index, which was measured in a maximum capacity test. Each subject then performed two cycling tests (at 60% and 70% maximum working capacity). The subject continued cycling until exhaustion in each test and then sat recovering until he/she no longer felt fatigue or until the oxygen uptake (VO2) and heart rate (HR) returned to their baselines, whichever was longer. The results indicated that HR required the longest time to recover and, consequently, HR data were adopted to set the CRT. The CRT was significantly correlated with the cardiorespiratory capability index and the relative workload indices: RVO2 and RHR. The RVO2 was the average elevation in VO2 during work from the resting level as a percentage of maximum VO2 reserve. The RHR's definition was similar to that of RVO2. Based on the obtained CRT-prediction model, the CRT for a high-cardiorespiratory-capability person was 20.8, 22.1, 23.4, and 24.7 min at 50%, 60%, 70%, and 80% RHR levels, respectively. These suggested CRT values should be increased by 10 min for a low-cardiorespiratory-capability person.
Journal of The Chinese Institute of Industrial Engineers, 2002
This study was aimed to establish a maximal oxygen uptake (VO 2max ) prediction model using non-e... more This study was aimed to establish a maximal oxygen uptake (VO 2max ) prediction model using non-exercise variables for young adults in Taiwan. Thirty young individuals participated in this experiment. The accurate VO 2max was measured using the K4 telemetry system during increment cycling test. A significant difference in VO 2max between males (3.34 ±0.69 liters min -1 ) and females (2.07 ±0.38 liters min -1 ) was found. The results indicate that VO 2max was significantly correlated with gender, age and body mass index (BMI). BMI (kg m -2 ) was computed from body weight in kilograms and body height in meters. The multi-regression model (N=24) for predicting VO 2max (liters min -1 ) was obtained as : VO 2max = 3.127 +(0.980 ×Gender) −(0.115 ×Age) +(0.084 ×BMI); R 2 =77.3%. The accuracy of the model was evaluated by conducting a cross-validation analysis (N=6).
European Journal of Applied Physiology, 2001
The aim of this study was to determine the maximum acceptable work duration (MAWD) for high-inten... more The aim of this study was to determine the maximum acceptable work duration (MAWD) for high-intensity work. Thirty young individuals participated in this study. Their maximum oxygen uptake ( \(\dot V{\rm O}_{{\rm 2max}} \) ) and maximum work rate (MWR) were assessed first. Each subject then performed two cycling tests (60% and 70% MWR) on two separate days. Oxygen uptake and heart rate data were collected throughout the test. The results indicate that the MAWD in the 60% MWR test (18.8 min) was about threefold greater than the MAWD in the 70% MWR test (6.5 min). The MAWD was inversely correlated with the relative workload indices: relative oxygen uptake ( \({\rm R}\dot V{\rm O}_{\rm 2} \) ; r=–0.82, P r=–0.79, P \({\rm R}\dot V{\rm O}_{\rm 2} \) was defined as the elevation in oxygen uptake from the resting level as a percentage of the difference between maximum and resting oxygen uptake. The RHR was defined as the elevation in heart rate from the resting level as a percentage of the difference between maximum and restingheart rate. Furthermore, more than 80% of the variations were explained by the exponential decrease regression model for predicting MAWD using the above two variables. The findings of this study can provide useful information for the design of high-intensity jobs.
Ergonomics, 2002
For safe job design, it is necessary to know maximum acceptable work time (MAWT) for a given work... more For safe job design, it is necessary to know maximum acceptable work time (MAWT) for a given workload. The aim of this study was to establish the relationship between MAWT and physical workload. Cycling tests at six different work rates relative to personal maximum working capacity were performed by 12 young adults in the laboratory. The oxygen uptake (VO 2 ) in percent of maximum aerobic capacity (%VO 2max ), relative heart rate (RHR), and relative oxygen uptake (RVO 2 ) were collected throughout the test.
Applied Soft Computing, 2009
ABSTRACT Estimating the cycle time of a job in a wafer fabrication plant (wafer fab) is a critica... more ABSTRACT Estimating the cycle time of a job in a wafer fabrication plant (wafer fab) is a critical task to the wafer fab. Many recent studies have shown that pre-classifying a job before estimating the cycle time was beneficial to the forecasting accuracy. However, most pre-classification approaches applied in this field could not absolutely classify jobs. Besides, whether the pre-classification approach combined with the subsequent forecasting approach was suitable for the data was questionable. For tackling these problems, two hybrid approaches with example post-classification, the equally-divided method and the proportional-to-error method, are proposed in this study in which a job is post-classified by a back propagation network (BPN) instead after the forecasting error is generated. In this novel way, only jobs whose cycle time forecasts are the same accurate will be clustered into the same category, and the classification algorithm becomes tailored to the forecasting approach. For evaluating the effectiveness of the proposed methodology and to make comparison with some existing approaches, production simulation (PS) is applied in this study to generate test data. According to experimental results, the forecasting accuracy (measured with root mean squared error, RMSE) of the proportional-to-error method was significantly better than those of the other approaches in most cases by achieving a 26–56% (and an average of 41%) reduction in RMSE over the comparison basis – multiple-factor linear combination (MFLC). The effect of post-classification was also statistically significant.
Output time prediction is a critical task to a wafer fab (fabrication plant). However, traditiona... more Output time prediction is a critical task to a wafer fab (fabrication plant). However, traditional wafer lot output time prediction methods are based on the historical data of the fab. The influence of the future release plan has been neglected. In addition, a lot that will be released in the future might appear in front of another lot that currently exists in the fab. For these reasons, to further improve the accuracy of wafer lot output time prediction, the future release plan of the fab has to be considered, and a hybrid ANN (SOM+FBPN) incorporating the future release plan of the fab is proposed in this study. Production simulation is also applied to generate test examples. According to experimental results, the prediction accuracy of the proposed methodology was significantly better than those of three approaches, FBPN, evolving fuzzy rules (EFR), and the hybrid ANN without considering the future release plan in most cases by achieving a 20%~49% (and an average of 35%) reduction in the root-mean-squared-error (RMSE) over the comparison basis -the FBPN.
Applied Ergonomics, 2010
This study utilized an external logger system for onsite measurements of computer activities of t... more This study utilized an external logger system for onsite measurements of computer activities of two professional groupsdtwelve university administrators and twelve computer-aided design (CAD) draftsmen. Computer use of each participant was recorded for 10 consecutive daysdan average of 7.9 AE 1.8 workdays and 7.8 AE 1.5 workdays for administrators and draftsmen, respectively. Quantitative parameters computed using recorded data were daily dynamic duration (DD) and static duration, daily keystrokes, mouse clicks, wheel scrolling counts, mouse movement and dragged distance, average typing and clicking rates, and average time holding down keys and mouse buttons. Significant group differences existed in the number of daily keystrokes (p < 0.0005) and mouse clicks (p < 0.0005), mouse distance moved (p < 0.0005), typing rate (p < 0.0001), daily mouse DD (p < 0.0001), and keyboard DD (p < 0.005). Both groups had significantly longer mouse DD than keyboard DD (p < 0.0001). Statistical analysis indicates that the duration of computer use for different computer tasks cannot be represented by a single formula with same set of quantitative parameters as those associated with mouse and keyboard activities. Results of this study demonstrate that computer exposure during different tasks cannot be estimated solely by computer use duration. Quantification of onsite computer activities is necessary when determining computer-associated risk of musculoskeletal disorders. Other significant findings are discussed.
International Journal of Industrial Ergonomics, 2009
This study was aimed at better understanding of the effects of ergonomics-based wafer-handling tr... more This study was aimed at better understanding of the effects of ergonomics-based wafer-handling training in a semiconductor fab. The 400 female wafer handlers were given 2 h of the ergonomics-based waferhandling training course. The identified risk factor ratio (IRFR) in the workplace and workers' musculoskeletal disorders (MSDs) were collected before and after the training. The qualitative efficacy of the training was also obtained through a questionnaire. The responses to the questionnaire reflect the generally positive attitude of the workers towards the training. The results of the IRFR from pre-to immediately post-training proved that the implementation of the training significantly increased safe behavior in work practices. One year after training, no significant decreases in the prevalence of MSDs were found for any body parts except the legs. This ergonomics-based training intervention is considered as a success to reduce risk factors associated with improper work methods and postures, but little data could validate its effectiveness on prevention of all the MSD problems. Relevance to industry: This study demonstrates a systematic approach for examining effects of ergonomics-based wafer-handling training tailored for female fab workers on reductions in work-related risk factors and their musculoskeletal disorders. Although the training course and effect evaluation methods are developed for the semiconductor industry, other similar industries can refer to this paper to design their own ergonomic training courses and evaluate the effects of the training interventions in practice.
International Journal of Industrial Ergonomics, 2009
This study presents a novel external device for quantifying accurately keyboard and mouse usage w... more This study presents a novel external device for quantifying accurately keyboard and mouse usage with a high input rate for prolonged periods. Three experimental tasks were performed to determine the effectiveness of the proposed logger. All mouse and keyboard activities were recorded simultaneously using the logger and WorkPace software. Ten university students typed on a keyboard, moved and clicked a mouse during tasks under various preset and self-selected speeds for 3 min. Computer input activities of 7 university students playing online games and 7 processing documents were recorded for 8 consecutive days. Analytical results demonstrate that the proposed external logger is more accurate than WorkPace, especially at a fast data entry speeds. The keystroke counts recorded by WorkPace were accurate only at low typing speeds. Moreover, the keystroke count recorded by WorkPace decreased by 17.6% from the logger-registered keystroke count when the typing rate was !4 keys/s, particularly when amalgamating with composite keystrokes. Similarly, the number of mouse clicks counted by WorkPace was 27.8% less than that counted by the logger when the click rate was !3.2 clicks/s. The average reduction in number of WorkPace-counted keystrokes and mouse clicks was 7.7 AE 2.0% (range [4.4%, 9.7%]) and 3.3 AE 3.3% (range [0.04%, 8.7%]), respectively, during online gaming tests, and was 4.0 AE 4.9% (range [À3.6%, 8.9%]) and 4.6 AE 4.7% (range [0.7%, 13.1%]), respectively, during document processing tests. This work demonstrates that the proposed external logger is an accurate and reliable tool for quantitatively measuring computer input activities. Further applications of the logger are discussed. Relevance to industry: Quantitative assessment of keyboard and mouse use can help identify high-risk groups and safe duration of computer use. The proposed logger effectively measures computer input activities, thereby providing accurate data than a software-based logger in assessing computer usage.
Ergonomics, 2011
This study explores the preferred viewing distance and character size for an electronic paper dis... more This study explores the preferred viewing distance and character size for an electronic paper display for three age groups. Proofreading speed and accuracy ratio were measured during Chinese proofreading tests using the preferred character size and minimum acceptable character size. Data analysis showed that the mean preferred viewing distance for young, middle-aged and older groups was 503, 455 and 444 mm, respectively. The mean preferred character size determined by young, middle-aged and older groups was 42.0, 50.0 and 55.2 min arc, respectively. The proofreading test results indicated that the older group proofread significantly more slowly (1.25word/sec) than the young (1.76 word/sec) and middle-aged groups (1.74 word/sec). Further, the participants proofread more correctly with their preferred character size (73.3%) than with their minimum acceptable character size (65.4%). This study provides valuable information for the design of Chinese text presentations for various agegroups. Statement of Relevance: This study confirmed the preferred viewing distance and character size for E-paper display were influenced by age. The preferred Chinese character size for young, middle-aged and older people was 42, 50 and 55 min arc, respectively. Therefore, the age factor should be considered for E-paper displays design and video display terminal (VDT) guidelines.
International Journal of Industrial Ergonomics, 2004
The semiconductor industry is moving from the production of 200 mm wafers to 300 mm wafers. With ... more The semiconductor industry is moving from the production of 200 mm wafers to 300 mm wafers. With the increase in wafer size, the workload of wafer handling tasks is also increasing. This study aims to evaluate the operator's handling capability and workload of handling 300 mm wafers by psychophysical, physiological and biomechanical approaches.
IEEE Transactions on Semiconductor Manufacturing, 2003
Semiconductor manufacturing industry is moving into the production of 300-mm wafers. To solve the... more Semiconductor manufacturing industry is moving into the production of 300-mm wafers. To solve the increased workload problem in manual wafer handling, some personal guided vehicles (PGVs) have been developed to help in the transfer of Front Opening Unified Pods (FOUP). This study compares two kinds of PGVs with a traditional cart and evaluates the feasibility of using them for manual FOUP handling tasks. Manual FOUP handling capability was assessed. The results indicate that there is no obvious advantage in using any of the two evaluated PGVs over the manual cart. There is potential risk of causing musculoskeletal disorders for female operators to handle the 300 mm FOUP manually. Since the development of a fully automated intra-bay FOUP handling system is a project of high technical difficulty, a combination of manual and automated handling is the current approach. To enhance the operator's health, safety and productivity, selection and training of operators, adequate design of handling tools and machine interface, assessment and balancing of workload are necessary.
Ergonomics, 2005
This study was aimed to investigate complete recovery time (CRT) after exhaustion in high-intensi... more This study was aimed to investigate complete recovery time (CRT) after exhaustion in high-intensity work. Twenty-four subjects were divided into two groups based on the cardiorespiratory capability index, which was measured in a maximum capacity test. Each subject then performed two cycling tests (at 60% and 70% maximum working capacity). The subject continued cycling until exhaustion in each test and then sat recovering until he/she no longer felt fatigue or until the oxygen uptake (VO2) and heart rate (HR) returned to their baselines, whichever was longer. The results indicated that HR required the longest time to recover and, consequently, HR data were adopted to set the CRT. The CRT was significantly correlated with the cardiorespiratory capability index and the relative workload indices: RVO2 and RHR. The RVO2 was the average elevation in VO2 during work from the resting level as a percentage of maximum VO2 reserve. The RHR's definition was similar to that of RVO2. Based on the obtained CRT-prediction model, the CRT for a high-cardiorespiratory-capability person was 20.8, 22.1, 23.4, and 24.7 min at 50%, 60%, 70%, and 80% RHR levels, respectively. These suggested CRT values should be increased by 10 min for a low-cardiorespiratory-capability person.
Journal of The Chinese Institute of Industrial Engineers, 2002
This study was aimed to establish a maximal oxygen uptake (VO 2max ) prediction model using non-e... more This study was aimed to establish a maximal oxygen uptake (VO 2max ) prediction model using non-exercise variables for young adults in Taiwan. Thirty young individuals participated in this experiment. The accurate VO 2max was measured using the K4 telemetry system during increment cycling test. A significant difference in VO 2max between males (3.34 ±0.69 liters min -1 ) and females (2.07 ±0.38 liters min -1 ) was found. The results indicate that VO 2max was significantly correlated with gender, age and body mass index (BMI). BMI (kg m -2 ) was computed from body weight in kilograms and body height in meters. The multi-regression model (N=24) for predicting VO 2max (liters min -1 ) was obtained as : VO 2max = 3.127 +(0.980 ×Gender) −(0.115 ×Age) +(0.084 ×BMI); R 2 =77.3%. The accuracy of the model was evaluated by conducting a cross-validation analysis (N=6).
European Journal of Applied Physiology, 2001
The aim of this study was to determine the maximum acceptable work duration (MAWD) for high-inten... more The aim of this study was to determine the maximum acceptable work duration (MAWD) for high-intensity work. Thirty young individuals participated in this study. Their maximum oxygen uptake ( \(\dot V{\rm O}_{{\rm 2max}} \) ) and maximum work rate (MWR) were assessed first. Each subject then performed two cycling tests (60% and 70% MWR) on two separate days. Oxygen uptake and heart rate data were collected throughout the test. The results indicate that the MAWD in the 60% MWR test (18.8 min) was about threefold greater than the MAWD in the 70% MWR test (6.5 min). The MAWD was inversely correlated with the relative workload indices: relative oxygen uptake ( \({\rm R}\dot V{\rm O}_{\rm 2} \) ; r=–0.82, P r=–0.79, P \({\rm R}\dot V{\rm O}_{\rm 2} \) was defined as the elevation in oxygen uptake from the resting level as a percentage of the difference between maximum and resting oxygen uptake. The RHR was defined as the elevation in heart rate from the resting level as a percentage of the difference between maximum and restingheart rate. Furthermore, more than 80% of the variations were explained by the exponential decrease regression model for predicting MAWD using the above two variables. The findings of this study can provide useful information for the design of high-intensity jobs.
Ergonomics, 2002
For safe job design, it is necessary to know maximum acceptable work time (MAWT) for a given work... more For safe job design, it is necessary to know maximum acceptable work time (MAWT) for a given workload. The aim of this study was to establish the relationship between MAWT and physical workload. Cycling tests at six different work rates relative to personal maximum working capacity were performed by 12 young adults in the laboratory. The oxygen uptake (VO 2 ) in percent of maximum aerobic capacity (%VO 2max ), relative heart rate (RHR), and relative oxygen uptake (RVO 2 ) were collected throughout the test.
Applied Soft Computing, 2009
ABSTRACT Estimating the cycle time of a job in a wafer fabrication plant (wafer fab) is a critica... more ABSTRACT Estimating the cycle time of a job in a wafer fabrication plant (wafer fab) is a critical task to the wafer fab. Many recent studies have shown that pre-classifying a job before estimating the cycle time was beneficial to the forecasting accuracy. However, most pre-classification approaches applied in this field could not absolutely classify jobs. Besides, whether the pre-classification approach combined with the subsequent forecasting approach was suitable for the data was questionable. For tackling these problems, two hybrid approaches with example post-classification, the equally-divided method and the proportional-to-error method, are proposed in this study in which a job is post-classified by a back propagation network (BPN) instead after the forecasting error is generated. In this novel way, only jobs whose cycle time forecasts are the same accurate will be clustered into the same category, and the classification algorithm becomes tailored to the forecasting approach. For evaluating the effectiveness of the proposed methodology and to make comparison with some existing approaches, production simulation (PS) is applied in this study to generate test data. According to experimental results, the forecasting accuracy (measured with root mean squared error, RMSE) of the proportional-to-error method was significantly better than those of the other approaches in most cases by achieving a 26–56% (and an average of 41%) reduction in RMSE over the comparison basis – multiple-factor linear combination (MFLC). The effect of post-classification was also statistically significant.
Output time prediction is a critical task to a wafer fab (fabrication plant). However, traditiona... more Output time prediction is a critical task to a wafer fab (fabrication plant). However, traditional wafer lot output time prediction methods are based on the historical data of the fab. The influence of the future release plan has been neglected. In addition, a lot that will be released in the future might appear in front of another lot that currently exists in the fab. For these reasons, to further improve the accuracy of wafer lot output time prediction, the future release plan of the fab has to be considered, and a hybrid ANN (SOM+FBPN) incorporating the future release plan of the fab is proposed in this study. Production simulation is also applied to generate test examples. According to experimental results, the prediction accuracy of the proposed methodology was significantly better than those of three approaches, FBPN, evolving fuzzy rules (EFR), and the hybrid ANN without considering the future release plan in most cases by achieving a 20%~49% (and an average of 35%) reduction in the root-mean-squared-error (RMSE) over the comparison basis -the FBPN.
Applied Ergonomics, 2010
This study utilized an external logger system for onsite measurements of computer activities of t... more This study utilized an external logger system for onsite measurements of computer activities of two professional groupsdtwelve university administrators and twelve computer-aided design (CAD) draftsmen. Computer use of each participant was recorded for 10 consecutive daysdan average of 7.9 AE 1.8 workdays and 7.8 AE 1.5 workdays for administrators and draftsmen, respectively. Quantitative parameters computed using recorded data were daily dynamic duration (DD) and static duration, daily keystrokes, mouse clicks, wheel scrolling counts, mouse movement and dragged distance, average typing and clicking rates, and average time holding down keys and mouse buttons. Significant group differences existed in the number of daily keystrokes (p < 0.0005) and mouse clicks (p < 0.0005), mouse distance moved (p < 0.0005), typing rate (p < 0.0001), daily mouse DD (p < 0.0001), and keyboard DD (p < 0.005). Both groups had significantly longer mouse DD than keyboard DD (p < 0.0001). Statistical analysis indicates that the duration of computer use for different computer tasks cannot be represented by a single formula with same set of quantitative parameters as those associated with mouse and keyboard activities. Results of this study demonstrate that computer exposure during different tasks cannot be estimated solely by computer use duration. Quantification of onsite computer activities is necessary when determining computer-associated risk of musculoskeletal disorders. Other significant findings are discussed.
International Journal of Industrial Ergonomics, 2009
This study was aimed at better understanding of the effects of ergonomics-based wafer-handling tr... more This study was aimed at better understanding of the effects of ergonomics-based wafer-handling training in a semiconductor fab. The 400 female wafer handlers were given 2 h of the ergonomics-based waferhandling training course. The identified risk factor ratio (IRFR) in the workplace and workers' musculoskeletal disorders (MSDs) were collected before and after the training. The qualitative efficacy of the training was also obtained through a questionnaire. The responses to the questionnaire reflect the generally positive attitude of the workers towards the training. The results of the IRFR from pre-to immediately post-training proved that the implementation of the training significantly increased safe behavior in work practices. One year after training, no significant decreases in the prevalence of MSDs were found for any body parts except the legs. This ergonomics-based training intervention is considered as a success to reduce risk factors associated with improper work methods and postures, but little data could validate its effectiveness on prevention of all the MSD problems. Relevance to industry: This study demonstrates a systematic approach for examining effects of ergonomics-based wafer-handling training tailored for female fab workers on reductions in work-related risk factors and their musculoskeletal disorders. Although the training course and effect evaluation methods are developed for the semiconductor industry, other similar industries can refer to this paper to design their own ergonomic training courses and evaluate the effects of the training interventions in practice.
International Journal of Industrial Ergonomics, 2009
This study presents a novel external device for quantifying accurately keyboard and mouse usage w... more This study presents a novel external device for quantifying accurately keyboard and mouse usage with a high input rate for prolonged periods. Three experimental tasks were performed to determine the effectiveness of the proposed logger. All mouse and keyboard activities were recorded simultaneously using the logger and WorkPace software. Ten university students typed on a keyboard, moved and clicked a mouse during tasks under various preset and self-selected speeds for 3 min. Computer input activities of 7 university students playing online games and 7 processing documents were recorded for 8 consecutive days. Analytical results demonstrate that the proposed external logger is more accurate than WorkPace, especially at a fast data entry speeds. The keystroke counts recorded by WorkPace were accurate only at low typing speeds. Moreover, the keystroke count recorded by WorkPace decreased by 17.6% from the logger-registered keystroke count when the typing rate was !4 keys/s, particularly when amalgamating with composite keystrokes. Similarly, the number of mouse clicks counted by WorkPace was 27.8% less than that counted by the logger when the click rate was !3.2 clicks/s. The average reduction in number of WorkPace-counted keystrokes and mouse clicks was 7.7 AE 2.0% (range [4.4%, 9.7%]) and 3.3 AE 3.3% (range [0.04%, 8.7%]), respectively, during online gaming tests, and was 4.0 AE 4.9% (range [À3.6%, 8.9%]) and 4.6 AE 4.7% (range [0.7%, 13.1%]), respectively, during document processing tests. This work demonstrates that the proposed external logger is an accurate and reliable tool for quantitatively measuring computer input activities. Further applications of the logger are discussed. Relevance to industry: Quantitative assessment of keyboard and mouse use can help identify high-risk groups and safe duration of computer use. The proposed logger effectively measures computer input activities, thereby providing accurate data than a software-based logger in assessing computer usage.