Dr. Joonwoo Son is a senior research scientist in the HumanLab at the Daegu Gyeongbuk Institute of Science and Technology (DGIST). He holds a B.S., an M.S. and Ph.D. from Pusan National University, South Korea. He has worked as a senior research engineer at Daewoo Precision Ind. Ltd. and SiemensVDO, and has been a visiting researcher at the Massachusetts Institute of Technology (MA, USA).
Dr. Son's primary research interests are in the areas of design for older drivers, driver¡¯s workload assessment methodology (on-road & simulator study), advanced human-vehicle interface, eco-drive assistant technologies. He has recently led research and development programs for agencies including Ministry of Knowledge Economy (MKE), Ministry of Education, Science and Technology (MEST), Ministry of Land, Transport and Maritime Affairs (MLTM) and Hyundai Motor Company, among others. He has authored or co-authored many conference and journal papers on a wide range of automotive engineering, human factors and transportation topics, and has consulted for various private companies and government agencies.
D e fin itio n o f u s a b ility c rite ria S e le c tio n o f e v a lu a tio n m e th o d s U sa... more D e fin itio n o f u s a b ility c rite ria S e le c tio n o f e v a lu a tio n m e th o d s U sa b ility C rite ria M et E x p e rim e n ta l m e th o d s O b je ctive m e th o d s L a te ra l d riv in g c o n tro l L o n g itu d in a l d riv in g c o n tro l V isu a l b e h a v io r / E ve n t d e te c tio n S e c o n d a r y ta sk tim e s S u b je c tive m e th o d s D A L I(S u b je c tiv e ra tin g s o f d riv e r w o rk lo a d) S U S (S u b je ctiv e ra tin g s o f sy ste m u sa b ility) Abstract-This paper proposes an in-vehicle interface assessment framework for emerging technologies. The framework was validated through a driving simulator based case study with an emerging user interface design. The result suggested that it was useful to evaluate the effectiveness of the emerging technology based user interface designs.
International Journal of Automotive Technology, 2021
Driver distraction is an important contributing factor to increase crash risks. The effect of sec... more Driver distraction is an important contributing factor to increase crash risks. The effect of secondary task engagement may vary by interaction types and driver age. Thus, it is essential to understand the impacts of distraction type and age on driving performance for improving the safety of in-vehicle secondary task design. This paper aims to assess potential risks induced by visual and auditory secondary tasks while driving. Thirty drivers, consisted of fifteen younger drivers aged 25 ∼ 35 and fifteen older aged 60 ∼ 69, were recruited and asked to drive in a simulator. They conducted two driving sessions, one for visually distracted driving and the other for cognitive distraction. The order in which secondary tasks were presented was counter-balanced. Driving performance and behaviour data were collected continuously using multiple measurement devices for vehicle speed, lane position, electrocardiogram, and gaze pattern. Differences in younger and older drivers’ performance while conducting the secondary tasks were compared. The result indicated that the effect of interaction types, i.e., visual and auditory, on older drivers’ performance was significant. More difficult secondary task creates greater age difference in driving performance. However, eye movement and physiological response were not significantly different between younger and older drivers. This result could suggest older drivers’ lower risk awareness of cognitive distraction.
Abstract—Recently, thanks to the rapid development of computer convergence technology, driver mon... more Abstract—Recently, thanks to the rapid development of computer convergence technology, driver monitoring system with various bionic sensor devices has become one of the most popular forms of technology to decrease the traffic accidents. In this paper, we present the issue of exploiting the driver's bionic data monitoring system by using ISO/IEEE 11073 and HL7 standard interfaces. The proposed system is based on the client and sever model, where the client agent encodes the health related data and server manager decodes the received message from the client. In order to gather the driver's health data, the client agent activates various bionic sensors which include not only personal social information but also height, weight, body temperature, SPO2, blood pressure and etc. Every sensor device is connected by configuring the wireless sensor network which is applicable to ZigBee or Bluetooth. After the health data is gathered in the remotely located database server, it can be ...
Objective: The aim of this study is to compare and evaluate the accuracy of workload distinction ... more Objective: The aim of this study is to compare and evaluate the accuracy of workload distinction of quantification measures (HR, SDNN, RMSSD, RMSE) by ROC analysis method to apply ECG biosignal to safe driving support technology of intelligent vehicle. And a system applied ROC analysis method to analyze the accuracy of workload distinction in real-time is developed. Background: Biosignals such as electrocardiography (ECG) can be used to analyze driver’s cognitive workload which is one of safe driving support technologies of the intelligent vehicle. Driver’s workload can be distinguished by analyzing heart rate (HR) and heart rate variability (HRV), which are calculated by ECG measured. To distinguish workload level based on ECG, individually suitable quantification measures of ECG should be selected; however, existing researches have used equal ECG quantification measures and ECG baseline. Method: System which can analyze workload in real-time and discriminate the most suitable meas...
Autonomous vehicles are promising for improving the safety and efficiency of transportation and m... more Autonomous vehicles are promising for improving the safety and efficiency of transportation and mobility systems by providing vehicle control during normal driving, or by providing emergency responses in safetycritical situations. Most major car makers announced that their autonomous vehicles will be available on the market by around 2020. However, a human driver will still be expected to have control responsibilities for an automated driving car in the case of emergency, particularly under NHTSA Levels 2 and 3 of automated driving. The driver’s role in an autonomous vehicle will be changed from a vehicle operator to a supervisor or an emergency operator, and the driver’s primary task will be to monitor critical factors in the environment or to detect system failures. Thus, shared-control paradigms in autonomous vehicles need to be addressed and discussed from the perspective of the driver’s changed role. The challenges include the control authority and transition between a human dr...
Stroke is a major cause leading to motor disability. Impaired motor function is one of the most s... more Stroke is a major cause leading to motor disability. Impaired motor function is one of the most serious causes of disabling sequelae of strokes, with over 50% of stroke patients experiencing a residual motor deficit. Researchers have shown an increased interest in proprioception deficits of the paretic side in hemiplegic patients. We examined the relationship between ankle proprioception and driving performance during simulator driving in post-stroke drivers. Four post-stroke drivers participated in this study. We developed an assessment environment using a driving simulator to evaluate driving performancce of stroke patients. The driving scenario consists of 3.5 km urban traffic conditions (3 minutes), 10 km divided 4-lane straight highway (6 minutes), and 7 km 2-lane curved or hilly rural roads (6 minutes). Performance parameters during the simulated drive were automatically generated by the simulator software. The parameters included time-to-collision (TTC), number of road edge e...
Objective: This paper aims to investigate age differences in the effectiveness of advanced warnin... more Objective: This paper aims to investigate age differences in the effectiveness of advanced warning systems. Background: Understanding of age differences in the effectiveness towards advanced driver assistance systems may help successful introduction of new technology in auto-market. Method: 52 drivers were participated in on-road field experiments. They drove approximately 6.2km on urban road (about 25 minutes) and 9.6km of highway (about 10 minutes) either with or without ADAS (Advanced Driver Assistance Systems) support. After completing the driving experiment, comparisons between the ADAS-supported and non-supported groups were conducted. Results: The results suggested that the FCW (Forward Collision Warning) and the LDW (Lane Departure Warning) significantly impact on average velocity and headway on highway driving, and steering reversal rate on urban driving, respectively. Conclusion: The ADAS are likely to make a positive influence on drivers’ behaviors by increasing their saf...
Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures f... more Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures for classifying a driver’s cognitive demand level from on-road experimental data. The measurement domains consist of driving performance, physiological arousal and eye behavioral change. Method: subjects (15 males in the 25-35 age range (M=27.9, SD=3.13)), experimental setup (an instrumented vehicle which consists of six video cameras, driving data logger, gaze tracker, and physiological measurement systems), procedure (20 minutes of driving exercise on a urban road and another 20 minutes of highway driving on 36km of highway), cognitive load (N-back task, an auditory delayed digit recall task was used to create periods of cognitive demand at three distinct levels), rating of driving workload (rating subjective driving workload after watching the experimental video clips by 4 different reviewers). Result: Potential measures of driver’s cognitive workload are suggested to estimate drivers’...
It is known that driving style is highly correlated with safety and fuel economy. In general, dri... more It is known that driving style is highly correlated with safety and fuel economy. In general, drivers who have moderate driving style shows safer driving behavior than aggressive drivers, and they may achieve higher fuel efficiency than their aggressive counterpart. Some previous studies found forward collision warning systems supported to change driving style toward safer behavior. However, their potential benefit of improving fuel economy was not clearly evaluated. This paper aims to analyze the effect of the advanced driver assistance systems (ADAS) on driving style and fuel economy using field operational test data from 52 participants. A half of participants were supported by ADAS and the others were not supported. Each participant drove approximately 5.5 km of rural road (about 10 minutes), 6.2 km of urban road (about 25 minutes) and 9.6 km of highway (about 10 minutes). The results suggested that there has been significant interaction between ADAS and gender on driving style,...
Objective: The aim of this study is to estimate driver’s aggressiveness using time headway of veh... more Objective: The aim of this study is to estimate driver’s aggressiveness using time headway of vehicle. Background: Driving style is one of the main factors that affect road safety. Especially, aggressive driving may cause car accidents with high possibility. Therefore, it is important to estimate driver’s aggressiveness to prevent car accidents. However existing methods calculating aggressiveness are too complicated and need different types of information such as aerodynamic coefficients. Method: 52 drivers (26 younger, 26 older) were participated in this experiment. They drove about 9.6km highway, and the driving data were synchronized and saved in same sampling rate. In this paper, more simple way to measure aggressiveness of driver was proposed by analyzing time headway and percent headway. Results: The result shows that the average time headway and the percent headway were highly correlated with the aggressiveness factor that was calculated using vehicle’s complex powers.
Recent technological advances have enabled a wide variety of information systems to be integrated... more Recent technological advances have enabled a wide variety of information systems to be integrated into a vehicle in order to increase safety, productivity, and comfort. However, improperly deployed technology can increase driver's workload and, consequently, degrade safety. Especially, potential information overload problems may become acute among older drivers who are the fastest growing segment of the driving population. Thus identification of a driver’s workload and spare capacity is crucial in the design of intelligent vehicles. With this knowledge, the in-vehicle information systems (IVIS) can provide timely and affordable information when the driver has the spare capacity to understand and respond it. This paper presents an empirical approach for estimating driver’s workload using driving performance, visual attention, and physiological indices. Moreover, the feasibity of diagnosticity to distinguish the type of driving workload was tested and a simple diagnositicity algor...
Objective: The aim of this study is to develop the upper limb training and assessment systems for... more Objective: The aim of this study is to develop the upper limb training and assessment systems for stroke patient rehabilitation using a driving simulator. Background: About 795,000 people experience a new or recurrent stroke in the United States each year, and lots of researchers are suggesting effective rehabilitation methods. In most of their studies, however, pre- and post-assessments tool using some clinical scales were used for the assessment tool of the recovery through upper limb rehabilitation. In some studies, physiological measure was considered as important factor for rehabilitation process from earlier studies. Method: The experiment for the assessment is conducted at a fixed-based driving simulator. The proposed rehabilitation assessment system consists of quantitative assessment using driving simulator, motion analysis system, EMG (Electromyograph), ECG (Electrocardiograph), EEG (Electroencephalograph), gaze tracker, body pressure sensor and myotonometer and subjective...
Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures f... more Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures for classifying a driver’s cognitive demand level from on-road experimental data. The measurement domains consist of driving performance, physiological arousal and eye behavioral change. Method: subjects (15 males in the 25-35 age range (M=27.9, SD=3.13)), experimental setup (an instrumented vehicle which consists of six video cameras, driving data logger, gaze tracker, and physiological measurement systems), procedure (20 minutes of driving exercise on a urban road and another 20 minutes of highway driving on 36km of highway), cognitive load (N-back task, an auditory delayed digit recall task was used to create periods of cognitive demand at three distinct levels), rating of driving workload (rating subjective driving workload after watching the experimental video clips by 4 different reviewers). Result: Potential measures of driver’s cognitive workload are suggested to estimate drivers’...
D e fin itio n o f u s a b ility c rite ria S e le c tio n o f e v a lu a tio n m e th o d s U sa... more D e fin itio n o f u s a b ility c rite ria S e le c tio n o f e v a lu a tio n m e th o d s U sa b ility C rite ria M et E x p e rim e n ta l m e th o d s O b je ctive m e th o d s L a te ra l d riv in g c o n tro l L o n g itu d in a l d riv in g c o n tro l V isu a l b e h a v io r / E ve n t d e te c tio n S e c o n d a r y ta sk tim e s S u b je c tive m e th o d s D A L I(S u b je c tiv e ra tin g s o f d riv e r w o rk lo a d) S U S (S u b je ctiv e ra tin g s o f sy ste m u sa b ility) Abstract-This paper proposes an in-vehicle interface assessment framework for emerging technologies. The framework was validated through a driving simulator based case study with an emerging user interface design. The result suggested that it was useful to evaluate the effectiveness of the emerging technology based user interface designs.
International Journal of Automotive Technology, 2021
Driver distraction is an important contributing factor to increase crash risks. The effect of sec... more Driver distraction is an important contributing factor to increase crash risks. The effect of secondary task engagement may vary by interaction types and driver age. Thus, it is essential to understand the impacts of distraction type and age on driving performance for improving the safety of in-vehicle secondary task design. This paper aims to assess potential risks induced by visual and auditory secondary tasks while driving. Thirty drivers, consisted of fifteen younger drivers aged 25 ∼ 35 and fifteen older aged 60 ∼ 69, were recruited and asked to drive in a simulator. They conducted two driving sessions, one for visually distracted driving and the other for cognitive distraction. The order in which secondary tasks were presented was counter-balanced. Driving performance and behaviour data were collected continuously using multiple measurement devices for vehicle speed, lane position, electrocardiogram, and gaze pattern. Differences in younger and older drivers’ performance while conducting the secondary tasks were compared. The result indicated that the effect of interaction types, i.e., visual and auditory, on older drivers’ performance was significant. More difficult secondary task creates greater age difference in driving performance. However, eye movement and physiological response were not significantly different between younger and older drivers. This result could suggest older drivers’ lower risk awareness of cognitive distraction.
Abstract—Recently, thanks to the rapid development of computer convergence technology, driver mon... more Abstract—Recently, thanks to the rapid development of computer convergence technology, driver monitoring system with various bionic sensor devices has become one of the most popular forms of technology to decrease the traffic accidents. In this paper, we present the issue of exploiting the driver's bionic data monitoring system by using ISO/IEEE 11073 and HL7 standard interfaces. The proposed system is based on the client and sever model, where the client agent encodes the health related data and server manager decodes the received message from the client. In order to gather the driver's health data, the client agent activates various bionic sensors which include not only personal social information but also height, weight, body temperature, SPO2, blood pressure and etc. Every sensor device is connected by configuring the wireless sensor network which is applicable to ZigBee or Bluetooth. After the health data is gathered in the remotely located database server, it can be ...
Objective: The aim of this study is to compare and evaluate the accuracy of workload distinction ... more Objective: The aim of this study is to compare and evaluate the accuracy of workload distinction of quantification measures (HR, SDNN, RMSSD, RMSE) by ROC analysis method to apply ECG biosignal to safe driving support technology of intelligent vehicle. And a system applied ROC analysis method to analyze the accuracy of workload distinction in real-time is developed. Background: Biosignals such as electrocardiography (ECG) can be used to analyze driver’s cognitive workload which is one of safe driving support technologies of the intelligent vehicle. Driver’s workload can be distinguished by analyzing heart rate (HR) and heart rate variability (HRV), which are calculated by ECG measured. To distinguish workload level based on ECG, individually suitable quantification measures of ECG should be selected; however, existing researches have used equal ECG quantification measures and ECG baseline. Method: System which can analyze workload in real-time and discriminate the most suitable meas...
Autonomous vehicles are promising for improving the safety and efficiency of transportation and m... more Autonomous vehicles are promising for improving the safety and efficiency of transportation and mobility systems by providing vehicle control during normal driving, or by providing emergency responses in safetycritical situations. Most major car makers announced that their autonomous vehicles will be available on the market by around 2020. However, a human driver will still be expected to have control responsibilities for an automated driving car in the case of emergency, particularly under NHTSA Levels 2 and 3 of automated driving. The driver’s role in an autonomous vehicle will be changed from a vehicle operator to a supervisor or an emergency operator, and the driver’s primary task will be to monitor critical factors in the environment or to detect system failures. Thus, shared-control paradigms in autonomous vehicles need to be addressed and discussed from the perspective of the driver’s changed role. The challenges include the control authority and transition between a human dr...
Stroke is a major cause leading to motor disability. Impaired motor function is one of the most s... more Stroke is a major cause leading to motor disability. Impaired motor function is one of the most serious causes of disabling sequelae of strokes, with over 50% of stroke patients experiencing a residual motor deficit. Researchers have shown an increased interest in proprioception deficits of the paretic side in hemiplegic patients. We examined the relationship between ankle proprioception and driving performance during simulator driving in post-stroke drivers. Four post-stroke drivers participated in this study. We developed an assessment environment using a driving simulator to evaluate driving performancce of stroke patients. The driving scenario consists of 3.5 km urban traffic conditions (3 minutes), 10 km divided 4-lane straight highway (6 minutes), and 7 km 2-lane curved or hilly rural roads (6 minutes). Performance parameters during the simulated drive were automatically generated by the simulator software. The parameters included time-to-collision (TTC), number of road edge e...
Objective: This paper aims to investigate age differences in the effectiveness of advanced warnin... more Objective: This paper aims to investigate age differences in the effectiveness of advanced warning systems. Background: Understanding of age differences in the effectiveness towards advanced driver assistance systems may help successful introduction of new technology in auto-market. Method: 52 drivers were participated in on-road field experiments. They drove approximately 6.2km on urban road (about 25 minutes) and 9.6km of highway (about 10 minutes) either with or without ADAS (Advanced Driver Assistance Systems) support. After completing the driving experiment, comparisons between the ADAS-supported and non-supported groups were conducted. Results: The results suggested that the FCW (Forward Collision Warning) and the LDW (Lane Departure Warning) significantly impact on average velocity and headway on highway driving, and steering reversal rate on urban driving, respectively. Conclusion: The ADAS are likely to make a positive influence on drivers’ behaviors by increasing their saf...
Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures f... more Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures for classifying a driver’s cognitive demand level from on-road experimental data. The measurement domains consist of driving performance, physiological arousal and eye behavioral change. Method: subjects (15 males in the 25-35 age range (M=27.9, SD=3.13)), experimental setup (an instrumented vehicle which consists of six video cameras, driving data logger, gaze tracker, and physiological measurement systems), procedure (20 minutes of driving exercise on a urban road and another 20 minutes of highway driving on 36km of highway), cognitive load (N-back task, an auditory delayed digit recall task was used to create periods of cognitive demand at three distinct levels), rating of driving workload (rating subjective driving workload after watching the experimental video clips by 4 different reviewers). Result: Potential measures of driver’s cognitive workload are suggested to estimate drivers’...
It is known that driving style is highly correlated with safety and fuel economy. In general, dri... more It is known that driving style is highly correlated with safety and fuel economy. In general, drivers who have moderate driving style shows safer driving behavior than aggressive drivers, and they may achieve higher fuel efficiency than their aggressive counterpart. Some previous studies found forward collision warning systems supported to change driving style toward safer behavior. However, their potential benefit of improving fuel economy was not clearly evaluated. This paper aims to analyze the effect of the advanced driver assistance systems (ADAS) on driving style and fuel economy using field operational test data from 52 participants. A half of participants were supported by ADAS and the others were not supported. Each participant drove approximately 5.5 km of rural road (about 10 minutes), 6.2 km of urban road (about 25 minutes) and 9.6 km of highway (about 10 minutes). The results suggested that there has been significant interaction between ADAS and gender on driving style,...
Objective: The aim of this study is to estimate driver’s aggressiveness using time headway of veh... more Objective: The aim of this study is to estimate driver’s aggressiveness using time headway of vehicle. Background: Driving style is one of the main factors that affect road safety. Especially, aggressive driving may cause car accidents with high possibility. Therefore, it is important to estimate driver’s aggressiveness to prevent car accidents. However existing methods calculating aggressiveness are too complicated and need different types of information such as aerodynamic coefficients. Method: 52 drivers (26 younger, 26 older) were participated in this experiment. They drove about 9.6km highway, and the driving data were synchronized and saved in same sampling rate. In this paper, more simple way to measure aggressiveness of driver was proposed by analyzing time headway and percent headway. Results: The result shows that the average time headway and the percent headway were highly correlated with the aggressiveness factor that was calculated using vehicle’s complex powers.
Recent technological advances have enabled a wide variety of information systems to be integrated... more Recent technological advances have enabled a wide variety of information systems to be integrated into a vehicle in order to increase safety, productivity, and comfort. However, improperly deployed technology can increase driver's workload and, consequently, degrade safety. Especially, potential information overload problems may become acute among older drivers who are the fastest growing segment of the driving population. Thus identification of a driver’s workload and spare capacity is crucial in the design of intelligent vehicles. With this knowledge, the in-vehicle information systems (IVIS) can provide timely and affordable information when the driver has the spare capacity to understand and respond it. This paper presents an empirical approach for estimating driver’s workload using driving performance, visual attention, and physiological indices. Moreover, the feasibity of diagnosticity to distinguish the type of driving workload was tested and a simple diagnositicity algor...
Objective: The aim of this study is to develop the upper limb training and assessment systems for... more Objective: The aim of this study is to develop the upper limb training and assessment systems for stroke patient rehabilitation using a driving simulator. Background: About 795,000 people experience a new or recurrent stroke in the United States each year, and lots of researchers are suggesting effective rehabilitation methods. In most of their studies, however, pre- and post-assessments tool using some clinical scales were used for the assessment tool of the recovery through upper limb rehabilitation. In some studies, physiological measure was considered as important factor for rehabilitation process from earlier studies. Method: The experiment for the assessment is conducted at a fixed-based driving simulator. The proposed rehabilitation assessment system consists of quantitative assessment using driving simulator, motion analysis system, EMG (Electromyograph), ECG (Electrocardiograph), EEG (Electroencephalograph), gaze tracker, body pressure sensor and myotonometer and subjective...
Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures f... more Objective: This paper aims to compare the sensitivity of multimodal cognitive workload measures for classifying a driver’s cognitive demand level from on-road experimental data. The measurement domains consist of driving performance, physiological arousal and eye behavioral change. Method: subjects (15 males in the 25-35 age range (M=27.9, SD=3.13)), experimental setup (an instrumented vehicle which consists of six video cameras, driving data logger, gaze tracker, and physiological measurement systems), procedure (20 minutes of driving exercise on a urban road and another 20 minutes of highway driving on 36km of highway), cognitive load (N-back task, an auditory delayed digit recall task was used to create periods of cognitive demand at three distinct levels), rating of driving workload (rating subjective driving workload after watching the experimental video clips by 4 different reviewers). Result: Potential measures of driver’s cognitive workload are suggested to estimate drivers’...
Uploads
Papers by Joonwoo Son