[2017.02] WEARABLE DESIGN FOR VIOLENT CRIME AGAINST CHILDREN

Page info

Name kidocee Date17-01-18 22:19 Hit482 Comment0

Contents

The aim of this project is to define problems and solution regarding violent crimes regarding children. The overall process begins with a preliminary research to gain holistic understanding of the design domains. This stage involved reading of literatures of studies and reports on the topic, gathering and analyzing of actual crime cases, and expert interviews. An integrated framework of sequence and experience of victim was created as the result. The researcher was also able to understand and grasp the characteristics and patterns of crimes against children. Children are significantly more vulnerable in their ability to recognize and respond to crimes, and often had hard time in giving accurate and adequate testimony. Criminals often take advantage of these vulnerabilities and that is why crime against children easily become habitual and concealed. A benchmark was conducted on the presently available solutions on the market as a part of the preliminary research. Most of the existing solutions were heavily dependent on the judgment and response of the user, which children fall short. Based on the findings of the preliminary research design goals were set. First goal is to aid response of the children and second is to aid recognition of the guardian. A police officer during her interview has stated that in 90 out of 100 cases, children under 12 would simply cry when they encounter crime. Therefore, the target user of this project was set to children of age 5 to 11, which includes children from pre-school to lower grade elementary school. Discovery of a set of data types is crucial in recognizing crime situation. A wearable device may utilize various types of sensors to gather this data set and translate them to notify the guardian regarding the child’s status. Some existing solutions in the market use circumstantial data as well as bio-signals since data from GPS and IMU is not sufficient to recognize desired conditions like seizure or baby crying. It is crucial to find an optimal set of data that is affective in detecting crime situation of children. Co-design method was used to attain this set of data. Stakeholders – Four experts and two mothers were invited to participate in a workshop that was designed to find the set of data that necessary for successful detection of crime against children and explore functional and morphological requirements of a device for specified purpose. As the result of the workshop, a set of data was concluded valuable for detection of crime against children: physiological signal (for emotion sensing), geographical location, movement and posture of the body. Translation of the readings on this set of data would trigger the alarm that would notify the guardian that there is a necessity to check on the child’s status through video and audio. Based on the result of the co-design workshop, traditional method of benchmarking, mood board, sketching and soft prototyping was conducted in order to find optimal design solution. Selection of the design was done with consultation of two professors and three graduate students. The product service system includes a wearable device worn by the child and a smart-device application for the guardian. The device constantly detects the location, posture, and bio-signal of the child. If there is change in the child’s status that is beyond the preset threshold, then an alarm is sent to the guardian, inducing a check-up. The guardian then checks the child’s status and makes assumptions based on the information of the location, posture, and emotion data translated from the bio signal. If the guardian judges the child may be in danger, then he or she is able to check the child’s status through making video call. Any necessary subsequent response can be made through the guardian’s discretion. Two working prototypes were manufactured: One for user validation and another for the proof of concept. Experiment for user validation was conducted with participation of total 11 eight-year-old children. The main goal of the experiment was to find out whether fear recognition using EDA and ECG is valid in everyday environment of children. Three types of data (EDA /ECG/ Acceleration in motion) were collected in four different conditions. (Calm, fear, excitement, physical exercise) The emotional stimulation of fear and excitement induced through the method used by Jang et al. (장은혜, 우태제, 이영창, & 손진훈, 2007) using the audiovisual film used to arouse fear. Both qualitative and quantitative analysis was conducted to check the validity of the concept based on the data set of three types: location, body movement, and emotion. The check the feasibility of recognizing emotion through physiological signals on daily level, an experiment was designed to check whether fear is distinguishable from other types of everyday stimulations – excitement and physical exercise. The visual inspection analysis showed that there is distinction between the EDA signal of fear and excitement. The repeated-measure ANOVA showed that there is correlation between the stimulation and stages in rate of change feather from EDA signals and all of the ECG signal features showed that all of the stimulations in the experiment was effective. Although there are definite advantages of the current concept such that the users will be able to respond to danger in faster based on the accurate data and plus many more, a clear limitation of this project is that it involves sophisticated data handling. The personal data collected needs to be interpreted and requires calibration to each individual user, which requires data science and deep learning technology. This project was an individual work of a designer, and product of such complexity usually requires collaboration of good number of team with its members from various expertise. The majority of human resource of start-up companies developing products of similar technology is composed of engineers and data scientists. Future work should be in a form of teamwork.

Comment list

No comment