通過可穿戴裝置最佳化以活動為基礎的嬰兒潮顾客參與度研究 - 政大學術集成

全文

(1)國立政治大學資訊管理學系. 碩士學位論文指導教授:苑守慈博士. 立. 政治大. ‧ 國. 學. ‧. 通過可穿戴裝置最佳化以活動為基礎的嬰兒潮顧客參與度研究 Optimizing Activity-Based Engagement of Baby Boomers through Wearable Devices. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. 研究生:魏曉晨中華民國 105 年 7 月. I.

(2) Optimizing Activity-Based Engagement of Baby Boomers through Wearable Devices 通過可穿戴裝置最佳化以活動為基礎的嬰兒潮顧客參與度研究運用錨定理論於創新服務特色之心理偏好分析 by Xiaochen Wei. A Dissertation Submitted in Partial Fulfillment of. 政治大. The Requirement for the Degree of Master of Science. 學. ‧ 國. 立. in Management information Systems. ‧ y. Nat. er. io. sit. Supervisor: Soe-Tysr Yuan, Professor, MIS, NCCU. al. n. iv n C h Management Department of i U Systems e n g c hInformation NATIONAL CHENGCHI UNIVERSITY July 2016 © Xiaochen Wei 2016. II.

(3) 謝辭. 台灣學習的兩年時間，不管是在學習上，還是在生活上，對我來說都是受益匪淺且終生難忘的回憶。首先要感謝我的指導教授－苑守慈老師，指引我在科研的道路上前行，鼓勵我在面對諸多困難時勇往直前，給予我像母親一般的照顧。正是因為老師的照料，才得以讓我完成這篇論文和完成研究所的學業。同時，我也時時刻刻被老師一絲不苟的治學態度，積極向上的生活態度所感染和影響。感謝謝沛宏博士和解燕豪教授，在論文口試期間，不吝給予的諸多寶貴建議，使本論文能夠更加嚴謹和完整。感謝王志臻機師，在論文的最初階段，不遠. 政治大無私與耐心的指導。感謝蔡焱昇學長，劉復漢學長，鄧光男叔叔，王桂蘭阿姨，立. 萬里，幫助我從美國買來設備和書籍。感謝黃裕福程式員，在論文實驗編程階段，. ‧ 國. 學. 在論文實驗階段，熱情地成為我們的受測者，配合我們進行繁瑣的實驗過程，給予我們重要的建議。. ‧. 感謝一共奮鬥的實驗室朋友－麥兜、庭毅、祖韻，兩年時間一起經歷過的討論，學習，鼓勵，幫助和嬉戲打鬧早已成為隨時想起來就能揚起嘴角的記憶。感. y. Nat. sit. 謝陪伴我成長的維正、志騰，小巫，文昕、建佑、柏緯、閎中，我們短暫的相處. n. al. er. io. 時光，依然充滿了各種歡樂的片段。感謝則匡學長，亦兄長，亦朋友般的關照。. i Un. v. 感謝政大資管所 103 級的各位伙伴，是你們讓我擁有了值得懷念的人與事。. Ch. engchi. 感謝我的父母，從精神和物質上支持我來台灣讀研究所的決定，支持我追求自己的夢想。感謝我的親人、朋友，因為你們，我才不會孤單。我將攜帶所有的愛，在未來的道路上努力前行。. III.

(4) 中文摘要. 現在，資訊技術為了通過實時、交互、有效的方式解決多樣但是具體的問題依然在發展變化著。然而幸運的是，以手機為代表的智慧產品已經發展到一定的階段，因此給了我們服務於不同的使用者提供了更多的機會。大約出生在第二次戰之後二十年的嬰兒潮一代，到現在處在 50-65 歲這一年齡階段。但是，相當數量的他們在經過幾十年的喧囂忙碌之後並沒有找尋到自己的生活意義。目前市場上的可穿戴裝置並不能滿足嬰兒潮人群的真正需求。在本研究中，我們想開發一個. 政治大期，監測期，優化期和自主期。之後資源將通過提供者和受益者之間的互動，達立通過可穿戴設備架構起來的服務交互平臺。此研究分為四個階段，分別是感應. 到“價值最大化”的最終目的。我們將要提供的服務，不僅部分預防了某些嚴重. ‧ 國. 學. 疾病的發生，同時也通過可穿戴裝置為嬰兒潮人群提供幫助，以實現有意義地生. ‧. Nat. y. 活。. n. al. er. io. 優化、介入。. sit. 關鍵詞：可穿戴裝置、嬰兒潮人群、行為識別、情緒識別、積極情緒、投入度，. Ch. engchi. IV. i Un. v.

(5) Abstract. Information technology nowadays is evolving and changing to solve various but specific human-needs problems through a real-time, interactive, effective way. Smart products exemplified by mobile phones have developed to reach a certain stage that gives us more chances to serve different users. Baby boomers, who were born in about twenty years after the second war, are now about 50-65 years old. A certain number of them confuse leisure in meaning of life after decades of bustle. However,. 政治大. current wearable devices still cannot fulfill the real demands of baby boomers. In this paper, we would like to develop interactive service system of wearable devices in four. 立. phases, respectively sensing, monitoring, optimization and autonomy, and then. ‧ 國. 學. resources are integrated through the interaction among providers and beneficiaries in order to reach the result of “value maximization” from each baby boomer's. ‧. perspective. This study aims to design services that not only prevent partly from serious diseases but also offer useful help to have a meaningful life through wearable. n. al. er. io. sit. y. Nat. devices.. Keywords:. Ch. engchi. i Un. v. Wearable device, baby boomer, activity recognition, emotion. recognition, positive emotion, engagement, optimization, intervention.. V.

(6) CONTENTS CHAPTER 1. INTRODUCTION ............................................................................... 1 1.1 Background and Motivation.............................................................................. 2 1.2 Research Question .............................................................................................. 3 1.3 Research Methods .............................................................................................. 6 1.4 Propose and contribution................................................................................... 9 1.5 Content Organization......................................................................................... 9 CHAPTER 2. LITERATURE REVIEW ................................................................. 12 2.1 Wearable Devices .............................................................................................. 12 2.1.1 The wearable devices - “Fitbit” and “Spire” .............................................. 14. 政治大 2.1.3 The Sensors of Emotional Data in Wearable Devices ................................ 16 立 2.2 The Service Design in Wearable Devices ........................................................ 16 2.1.2 The Sensors of Physical Data in Wearable Devices.................................... 15. ‧ 國. 學. CHAPTER 3. iEnjOrange PROJECT ..................................................................... 19 3.1 The Conceptual Framework of iEnjOrange .................................................. 19. ‧. 3.2 The Ecosystem of iEnjOrange ......................................................................... 21. sit. y. Nat. 3.3 The System Architecture of iEnjOrange ........................................................ 23. io. er. 3.4 The System Flow ............................................................................................... 25 CHAPTER 4. THE ACTIVITY-BASED ENGAGEMENT OF BABYBOOMERS. n. al. Ch. i Un. v. MECHANISM THROUGH WEARABLE DEVICES ........................................... 29. engchi. 4.1 Conceptual Framework ................................................................................... 29 4.2 Design logic and research approach ............................................................... 31 4.3 Activity-based Sensing Module ....................................................................... 33 4.4 Activity-based Monitor Module ...................................................................... 35 4.4.1 Activity ............................................................................................................ 35 4.4.2 Emotion .......................................................................................................... 38 4.4.3 Others ............................................................................................................. 39 4.5 Activity-based Optimization Module .............................................................. 40 4.6Activity-based Autonomy Module ................................................................... 43 CHAPTER 5. APPLICATION SCENARIO ........................................................... 46 CHAPTER 6 EVALUATION .................................................................................... 50 VI.

(7) 6.1 Propositions ....................................................................................................... 50 6.1.1 Assumptions ................................................................................................... 50 6.2 Controlled Experiment and Interview for Baby Boomers ........................... 51 6.2.2 Subjects of Controlled Experiment with Baby Boomers ........................... 53 6.2.3 Results of Controlled Experiment with Baby Boomers ............................. 54 6.3 Discussion of Findings ...................................................................................... 72 6.3.1 Propositions .................................................................................................... 73 CHAPTER 7. CONCLUSIONS................................................................................ 76 7.1 Contribution...................................................................................................... 76 7.2 Managerial Implications .................................................................................. 77 7.3 Limitations and Future Works ........................................................................ 77. 政治大. REFERENCES ........................................................................................................... 79. 立. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. VII. i Un. v.

(8) LIST OF FIGURES Figure1.1 Mobile phone use by age groups (Pew Internet Research report 2012) 3 Figure1.2 Service Design Model ................................................................................. 4 Figure 1.3 Four Phases in Service Interaction System of Wearable Device ........... 6 Figure 1.4 iEnjOrange Functions ............................................................................... 8 Figure 1.5 Positive Emotion Function ........................................................................ 8 Figure 1.6 Research Procedures ............................................................................... 11 Figure 2.1 Unit Operations in Obtaining Situational Awareness (Park et.al, 2014) ............................................................................................................................... 13. 政治大. Figure 3.2. Conceptual Building Block .................................................................. 19. Figure 3.3. The iEnjOrange Ecosystem ................................................................. 22. 立. Figure 3.4 The iEnjOrange System Architecture .................................................... 24. ‧ 國. 學. Figure 4.1 Conceptual Framework Diagram .......................................................... 29 Figure 4.2 System Architecture ................................................................................. 31. ‧. Figure 4.3 Body Sensor Networks (Sazonov et.al, 2014) ........................................ 33 Figure 4.4 Fitbit Surge’s Appearance & Fitbit’s Wearing Style ............................ 35. Nat. sit. y. Figure 4.5 (a) Spire’s Appearance & Spire’s Wearing Style .................................. 35. al. er. io. Figure 4.6 Recognized Activity Patterns .................................................................. 36. v. n. Figure 4.7 The Heart Rate Changing of One Person in The Whole Day .............. 36. Ch. i Un. Figure 4.8 The Posture Changes in The Sleep Condition ....................................... 37. engchi. Figure 4.9 The Relationship between Respiration and Emotion ........................... 38 Figure 4.10 The Example of Emotion Measurement in Spire ................................ 39 Figure 5.1 The Scores of PERMA ............................................................................. 47 Figure 5.2 The Recommended Activities.................................................................. 48 Figure 5.3 Providers in the System ........................................................................... 49 Figure 6.2.1 The Procedure of Controlled Experiment for Baby Boomers .......... 51 Figure 6.2.2 The Activity’s Scores of Subject A ....................................................... 60 Figure 6.2.3 The Emotion’s Scores of Subject A ..................................................... 60 Figure 6.2.4 The Activity’s Scores of Subject B ....................................................... 61 Figure 6.2.5 The Emotion’s Scores of Subject B ..................................................... 61 Figure 6.2.6 The Activity’s Scores of Subject C ...................................................... 62 VIII.

(9) Figure 6.2.7 The Emotion’s Scores of Subject C ..................................................... 62 Figure 6.2.8 The Activity’s Scores of Subject D ...................................................... 63 Figure 6.2.9 The Emotion’s Scores of Subject D ..................................................... 63 Figure 6.2.9 The Heart Rate in A Day of Subject A ................................................ 65 Figure 6.2.10 The Steps in A Day of Subject A ........................................................ 65 Figure 6.2.11 The Breath Rate in A Day of Subject A............................................. 66 Figure 6.2.12 The Ratio of Actual Value to Expected Value ................................... 69 Figure 6.3.1 Propositions ........................................................................................... 73. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. IX. i Un. v.

(10) LIST OF TABLES Table 1.1 iEnjOrange Sensor Data ............................................................................. 7 Table 2.1 The available wearable devices in the market ........................................ 13 Table 2.2 Possible Parameters for Wearable Inertial Sensors ............................... 15 Table 3.1 Context Tags in Three Perspectives ......................................................... 27 Table 4.1 The Relationship between Basic Functions and Sensors ....................... 34 Table 4.2 Activity Engagement Weight .................................................................... 42 Table 4.3 Activity Engagement n .............................................................................. 42 Table 4.4 An Original Example of User Scenario ................................................... 43. 政治大. Table 4.5 The Engagement and Positive Emotion Scores of the User in A week .. 44 Table 5.1 An Ordinary Schedule of Sam On the Workday .................................... 46. 立. Table 6.2.1 The Survey of Setting Goal and Weight ............................................... 52. ‧ 國. 學. Table 6.2.2 The Background of Subjects .................................................................. 53 Table 6.2.3 Findings from the Interviews of Baby Boomers .................................. 55. ‧. Table 6.2.4 The Data Pattern from Wearable Devices ............................................ 56 Table 6.2.5 The Actual Situation of Data Pattern Obtainment .............................. 57. Nat. sit. y. Table 6.2.6 The Setting of Subject A ......................................................................... 58. io. er. Table 6.2.7 The Setting of Subject B ......................................................................... 58 Table 6.2.8 The Setting of Subject C ........................................................................ 58. n. al. Ch. i Un. v. Table 6.2.9 The Setting of Subject D ........................................................................ 59. engchi. Table 6.2.10 The Result in A Day of Subject A ........................................................ 66 Table 6.2.11 The Engagement Score of Subject A ................................................... 67 Table 6.2.12 The Positive Emotion Score of Subject A ........................................... 67 Table 6.2.13 The Engagement Score of Subject B ................................................... 67 Table 6.2.14 The Positive Emotion Score of Subject B ........................................... 68 Table 6.2.15 The Engagement Score of Subject C................................................... 68 Table 6.2.16 The Positive Emotion Score of Subject C ........................................... 68 Table 6.2.15 The Engagement Score of Subject D................................................... 68 Table 6.2.16 The Positive Emotion Score of Subject D ........................................... 69. 1.

(11) CHAPTER 1. INTRODUCTION 1.1 Background and Motivation. Now information technology is still changing in order to solve various but specific problems through a real-time, interactive, effective way. Fortunately, smart products represented by mobile phones have developed to reach a certain stage that gives us more chances to serve different users. Information technology has a close relationship with product itself: sensors, processors, software, and connectivity of products, associated cloud space for storing product data, and some applications for showing and analyzing data (Michael Porter, 2015). Wearable device as a new smart product owns an absolute advantage to fulfill above mentioned both hardware and software. 政治大 Baby boomers, who were born in about twenty years after the second war, are 立. functions in portability and in continuity.. about 50-65 years old now. Perhaps some of them have already retired, and some of. ‧ 國. 學. them still are on working. Perhaps some of them are trying to enjoy life, but some of them have no idea about it. However, they all need to plan their future lives that are. ‧. totally different from the previous. Meanwhile, health problems have begun. complex situation that needs us to pay special attention to.. io. sit. y. Nat. emerging, although most will not threat their lives. These people are facing such a. er. Is there any connection between wearable device and baby boomer? Most. al. n. iv n C h e n isg cfocused technology trends. The kind of product h i Uon entertainment; however, its. wearable devices (ex. Apple watch) are designed for the young who love pursuing. practical function is limited. The rest are designed for the old or patients who cannot look after themselves or who are sick seriously. This kind of product is simply focused on health caring. As indicated in Figure 1.1, the number of old people who use mobile technology, no matter it is social network, text message, use of the Internet, or use of mobile phone/tablets, is increasing. We suppose to offer services that not only prevent partly from serious diseases but also offer help useful help to have a meaningful life through wearable devices.. 2.

(12) 90%. 80% 70% 60% 50%. Apps Downloads. 40%. Text Messaging. 30%. Mobile e-mail. 20%. Access Internet. 10% 0% Younger Bommers (47 -56). Older Bommers (57 -65). Silent G.I Generation Generation (66 -74) (75+). 治政 Figure1.1 Mobile phone use by age groups (Pew Internet 大 Research report 2012) 立 ‧ 國. 學. 1.2 Research Question. ‧. In the traditional thinking, the core is product itself that means value exchanges. y. Nat. through specific product (Vargo, Lusch, 2004; Vargo, Morgan, 2005). From this view,. sit. the roles of “producer” and “consumer” are distinct, and the producer creates value. er. io. only before the consumer really uses the product (Stephen Vargo, Paul Maglio,. al. iv n C But in our research, wehfirstly pay attention e n g c h i U to service interactive system; n. Melissa Akaka, 2008). So the functions of product are not based on consumers’ actual demands.. in other words, value here is co-created by the producer and the consumer. Then resources are integrated through the interaction among providers and beneficiaries in order to reach the result of “value maximization”. From this view, the service is most important part in our total design process. So, we will design our wearable device from the view of service. Then we will explain details about our service design (see Figure 1.2). We aim at building a circle ecosystem that will translate hardware “sensor” into software “service”.. 3.

(13) Service 50~65 characteristics. PROVIDER. Data Support Specific Service. Platform Analysis & Store Data. 立. 政Sensor 治大. USER. Collect data. ‧. ‧ 國. 學 y. sit. io. There are three key points including in this model:. n. al. er. Nat. Figure1.2 Service Design Model. i Un. v. a) The appropriate methodology to identify baby boomers’ actual expectation.. Ch. engchi. b) The appropriate product to balance the difference between expectation and limitation. c) The appropriate service to deliver from producer to baby boomers through the product. We consider these key points from following four aspects: a). SERVICE The people who we call “Baby Boomer” generation need is quite different. They. do not need to pay much attention to their grown children so that they have much more free time than before. They have the last chance to grasp time to enjoy their lives before becoming aged. The service should be designed specially when taking the age characteristics into account, such as the part of health, which mainly consider from the respect of preventive medicine. In the review of James Prochaska (2008). 4.

(14) about the future of preventive medicine, multiple behaviors of individuals have a close relationship with chronic disease, disability, and premature death. He also points that tailored intervention that is based on the Internet technology are more effective in treatments, because it is interactive regardless of time and locations. We intend to design our services that are based on general activities, but are specific to users’ own conditions with different values, for the purpose of prevent serious disease through changing and keeping daily health activities simply. b) DATA Both provider and users design every specific service together, which is a process of “co-creation”. When decide to what kind of service to offer, we will think about “DATA”. Different kinds of data through group can implement different. 政治大. functions. The part of “data collection” is completed by users, and the part of “data analysis” is processed by service provider. In this research, service is essential, while. 立. data is elements of service. These data help us build a virtual user through various. ‧ 國. 學. numbers, because every specific data value is on behalf of an implicated message of the user’s life condition.. ‧. c) SENSOR. The traditional body sensors include vital sign monitors, activity monitors and. Nat. sit. y. sleep monitors – that are now also connected to the Internet. In addition, a new sensor. al. er. io. that can monitor emotion has been invented now. Users though the sensors they will. n. transmit their data to cellphone or computer with the technology of blue tooth or. Ch. i Un. v. wireless. In this way, the analog signals of body are not only conversed to digital. engchi. signals, but also recorded by memory as data that are used to analyze at once or later. d) FEEDBACK According to previous data, service will change automatically in order to adjust every specific user’s actual needs. Then efficient suggestions are offered to their users and can be felt through their wearable devices. The key point in this part is pattern and timing of intervention, that is when and how suggestions are delivered to users, for the purpose of building a service interactive system that can communicate with users based on their actual needs, just like their intimate friend who know about users better than themselves. However, most available wearable devices in the markets are physiological activities monitor, which only try to record and show the data. We probably have a. 5.

(15) question that what else can we do after knowing our data of body. Obviously, we have more expectations on this “little friend”, so we intend to make it friendlier.. The research question in this study as followings: 1. To explore how to design the activity-based mechanism to aid the baby boomers on life-style from present to long term. 2. To explore how to optimize the functions to combine the physical health and mental health under the automatic condition. 3. To explore how to utilize the intervention to have a comprehensive health to achieve the goal of “value-maximization” through sensors controlled by existing technology.. 1.3 Research Methods. 立. 政治大. In this paper, we would like to develop service interactive system of wearable device. ‧ 國. 學. in four phases, according to Michael Porter and James Heppelmann (2014)’s view about smart, connected products, respectively are sensor, control, optimization and. ‧. autonomy (see Figure 1.3). Here, we take the project of “iEnjOrange” as an example.. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. Figure 1.3 Four Phases in Service Interaction System of Wearable Device. a). SENSOR This phase mainly considers about data read and data access.. We will use the. sensors of existing products instead of designing own product. Two products will help us complete the whole research, and some improvements will be added to existing. 6.

(16) API. In addition, the balance between usability and endurance will be taken into account. Fitbit, which is a new smart sports watch, mainly has features - track and record all-day activities and sleeping conditions, and coordinate with smart phone and computers to show the results. Spire, which is an emotion monitor, has mainly feature - track breath to discover which condition – clam, focus or tense users in. We will combine these two products to achieve our goal of service interactive system, the data we want to read from sensor as Table1.1. Table 1.1 iEnjOrange Sensor Data HEALTH ACTIVITY RESPIRATION. BASIC LOCATION. Heart rate. Activity pattern. Breath rate. 治 goal 政 Activity 大. TIME. Blood pressure. GRAVITY. 立. Deep breath Breath pattern. ‧ 國. 學. b) MONITORING. ‧. This phase mainly considers about combining sensors and external data sources, such like social data, then arrangement of data in order to realize the functions. Nat. sit. y. through remote commands or algorithms (see Figure 1.4). What is more, this phase. io. er. focuses on the current status.. We will change Emotion Recognition API in order to improve user experience.. n. al. Ch. i Un. v. The emotion of tense, clam and focus can be notified in Spire, but there is possibility. engchi. to recognition more patterns of emotion when we have data of respiration. Specific patterns of respiration have a close relationship with human basic emotions.. 7.

(17) Calories/Glucose. Health Heart rate. Sleep. Sport. Health Heart rate. Basic. Emotion. Respiration. Nutrition. Activity Basic Health Heart rate. Basic Respiration. Sport. Health Heart rate. Respiration. 政治大 Figure 立 1.4 iEnjOrange Functions. ‧ 國. 學. c) OPTIMAZATION. ‧. This phase, which includes two aspects - positive emotion and engagement,. y. Nat. mainly considers about optimize the algorithms in order to give users more effective. sit. suggestions. What is more, this phase focuses on long-time status.. n. al. er. io. 1) Positive Emotion (see Figure 1.5). Nutrition. Sleep. Ch. eSport ngchi. iv n U Health. Emotion Spire. Fitbit. Figure 1.5 Positive Emotion Function. Here, the two products will be combined in order to measure both physical and mental situation when users are active. 2) Engagement This aims to measure mental state, which includes “Focus”, “Clam” and “Tension”, when you are doing a certain thing. Two dimensions – time and activity, from two aspects – frequency and degree, will analyze the result of short term.. 8.

(18) We will classify various activities into several patterns according to their normal respiratory status. Every pattern will have its own preset ratio and weight value. When a user completes an activity, there is a computed ratio. Then the actual ratio is compared with the preset ratio that gets a sore from 0 to 10. We can sum or average the sores by activity, and therefore get the engagement of a certain activity. Or we can sum or average the sore * the weight value in order to get the engagement by a certain time. In additions, a key problem here is intervention. Although we get the result of engagement, we still need to predict next stage and find the timing to encourage or stop.. d) AUTONOMY. 政治大. This phase mainly considers about adjustment in order to help users live a. 立. comprehensive health life according to the feedback from wearable devices.. ‧ 國. 學. We will give suggestion at first, and then the quality of completing will have an important effect on users’ next stage of suggestion. In this process, there are two. ecosystem.. ‧. patterns that are incentive and command as additions to enrich the scope of whole. y. Nat. er. io. al. sit. 1.4 Propose and contribution. n. The real-time data are recorded by sensors, are transmitted to smart phone or Internet,. Ch. and then are processed by platform successively.. engchi. i Un. v. Under such a condition, we. propose a mechanism that will not only use appreciable physiological data to measure comprehensive health, but also predict from current to future and advice users in time. What is more, the several third parties will be invented to join in with us in order to offer more abundant services to improve user experience. We also would like to build a structure of service interactive system cooperate with different areas of service provider (i.e., insurance company and hospital and etc.), offer integrate services. Besides, we will regard the ecosystem perspective that can improve the value co-creation with different actors. By the side of users, they are possible to have a high-quality life, and at last reach the goal of meaningful life.. 1.5 Content Organization. 9.

(19) The paper is organized as follows (see Figure 1.6). The next chapter reviews the literature on wearable sensors or devices, the relationship between respiration and emotion, positive emotion, engagement, customer needs about baby boomers to help us realize specific service methods. In Chapter 3, we will have a picture of motivating applications through building conceptual framework, ecosystem and system architecture of iEnjOrange. Then in Chapter 4, a specific mechanism will be come up with based on the Chapter 2 and tried to establish architecture with information technology to develop a possible solution that can be realized. We give an application scenario in Chapter 5 to present how our system works. At last, a conclusion and future work will be shown.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 10. i Un. v.

(20) CHAPTER 1 INTRODUCTION Background, Motivation, Question, Method, Purpose, Contribution, and Content Organization. CHAPTER 2 LITERATURE REVIEW Wearable sensors or devices, the relationship between respiration and emotion, positive emotion, engagement, customer needs about baby boomers. 政治大. CHAPTER 3 MOTIVATING APPLICATIONS Conceptual framework, ecosystem and system architecture of iEnOrange. 立. ‧ 國. 學 ‧. CHAPTER 4 MECHANISM Motivating Concept, Design Logic and Research Approach and Control Modules. n. er. io. sit. y. Nat. al. i Un. v. CHAPTER 5 APPLICATION SCENARIO An Example to Present How the System Works. Ch. engchi. CHAPTER 6 ENVALUATION The Design, Result and Discussion of the Experiment. CHAPTER 7 CONCLUSIONS Conclusion, Managerial Implication, Future Works Figure 1.6 Research Procedures. 11.

(21) CHAPTER 2. LITERATURE REVIEW In this section, the foundations and methodologies as the background knowledge of our research are discussed. The review of wearable devices, activity recognition in wearable devices, emotion recognition in wearable devices, and the service design of baby boomers are elaborated in order to build the fundamental as the required to support the research.. 2.1 Wearable Devices Wearable device, which is described as a new prevalent style of smart products, is mainly connected to the smart phone. These two things with the technology of big. 政治大 capabilities of smart products can be grouped into four areas: monitoring, control, 立. data and cloud computing together fulfill the integrated functions for users. The. optimization, and autonomy, and each builds on the preceding one (Porter &. ‧ 國. 學. Heppelmann, 2015). They say that monitoring comprehensively consists of the product’s condition, the external environment, the product’s operation and usage;. ‧. control personally consists of the product’s functions and user experience;. y. Nat. optimization is a stage that algorithms can enhance product performance and allow autonomous product operation,. io. sit. predictive service; autonomy is a stage that allows. er. self-coordination of operation with other products and systems and autonomous. al. n. Wearable device i v is aimed to find more n C valuable information in physical h data i Uof sensors. As they mention, e nwith g ctheh help product enhancement and personalization.. wearable devices can offer and create much more value in the digital environment. As Park et.al (2014) review, wearables perform five basic functions or unit operations in each scenario, respectively are sense, process (analyze), transmit, apply (utilize), especially processing may occur at a remote location (see as Figure 2.1). A sensor is defined as “a device used to detect, locate, or quantify energy or matter, giving a signal for the detection of a physical or chemical property to which the device responds” (Kress-Roger, 1997). But not all sensors can be easily worn, so there are eight key attributes in an ideal wearable device. On the one hand, from the view of physics, it should be lightweight, aesthetically pleasing, invisible and shape conformable. On the other hand, from the view of function, it should be multi-functional, configurable, responsive and bandwidth. In our research, we use two. 12.

(22) available products – “Fitbit” and “Spire”, to design our services.. 治政 Figure 2.1 Unit Operations in Obtaining Situational大 Awareness (Park et.al, 2014) 立 ‧ 國. 學. We collect the information about available devices in the market, and summary them in Table 2.1.. SPIRE. Surge Watch. WASHER-PROOF SENSOR. XIAOMI. vivosmart Watch. Stone. al. Wristband. i v7 days n 7 days C h 1day 7 days engchi U. n OPERATION TIME. GARMIN. y. APPLE. sit. io. TYPE. Nat. MODEL. FITBIT. er. BRAND. ‧. Table 2.1 The available wearable devices in the market. YES. YES. Wristban d 30 days. Heart rate. YES. YES. YES. YES *1. 3-aixs. YES. YES. YES. YES. YES. YES. YES. YES. accelerom eter/ gyro GPS. YES. WIRELESS. Bluetooth. YES. YES. FUNCTION. Activity. YES. YES. YES. YES *2. YES. Tracker Sleep Condition. 13. YES.

(23) Calories VIBRATOR. YES. YES. YES. YES. YES *3. YES. YES. YES. MICROPHONE Note:. *1&3 :This function need to buy additional accessories; *2 : This function need to download additional apps from App Store.）. 2.1.1 The wearable devices - “Fitbit” and “Spire”. Fitbit Surge, which is a new and popular smart sports watch, has the following features (see from www.fitbit.com/ surge): . split times. 立. PurePulseTM Heart Rate: Get continuous, automatic, wrist-based heart rate and. 學. ‧ 國. . 政治大. GPS Tracking: see distance, pace and elevation climbed and review routes and. simplified heart rate zones. All-Day Activity: Track steps, distance, calories burned, floor climbed and active. . ‧. minutes. Multi- Sport: Record running, cycling, cross training and other workouts and. Nat. y. . er. Long Battery Life: Lasts longer than competing trackers with a battery life up to. io. . sit. view exercise summaries. al. . Auto Sleep + Alarms: Monitor your sleep automatically and set a silent alarm. . Wireless Syncing: Sync stats wirelessly and automatically computers. n. . iv n C Notification + Music: See callh and text notifications e n g c h i U on display and control songs from your mobile playlist seven day. Spire, which is an emotion monitor, has following features (see from www.spire.io): . Track every step and every breath. . Be notified when users are tense or have not taken a deep breath. . Discover what makes you clam and focused. . Advanced step and calorie tracking. . Guided breathing exercises and meditations These two products have satisfied the basic eight key attributes we mention. 14.

(24) before, what we do in this research is to to integrate and improve the functions that can offer better services for users. 2.1.2 The Sensors of Physical Data in Wearable Devices. Inertial sensors, including accelerometers, gyroscopic sensors, and magnetic sensors, have small size, for monitoring the motion associated with human activities. Accelerometer (triaxial accelerometer is common) is used to measure motion; gyroscopic sensor is used to measure angular velocity; magnetic sensor is used to precisely measure body movement (Yang & Hsu 2010). The possible parameters for wearable inertial sensors are shown in the Table 2.2 (Tamura, 2014). He also mentions. 政治大. that static and dynamic activities can be measured through accelerometry; however, normal and abnormal walking or aged-related functions should be measured through. 立. amplitude and RMS values.. ‧ 國. 學. Table 2.2 Possible Parameters for Wearable Inertial Sensors. ‧. PARAMETERS. Triaxial acceleration. y. Nat. VELOCITY. SIGNALS. NUMBER OF STRIDES. Triaxial acceleration. n. al. er. sit. Triaxial acceleration. io. NUMBER OF STEPS. TIME. AVERAGE STRIDE TIME. Ch. Triaxial acceleration. CADENCE ROOT. MEAN. iv n U e n gTriaxial c h i acceleration Triaxial acceleration. SQUARE. (RMS) Triaxial acceleration, triaxial gyroscopic. VALUES. velocity & triaxial magnetic field. REGULARITY OF STEPS. Triaxial acceleration. REGULARITY OF STRIDES. Triaxial acceleration. ANGLE. Combined with triaxial acceleration, triaxial angular velocity & triaxial magnetic field. Another common kind of values is heart rate, which can be got from assessing the arterial pulsatility skin vascular beds with LED technology, which is called. 15.

(25) Photoplethysmography (PPG). According to Lemay et.al (2014), it illuminates a living tissue with a light beam, and capture a portion of the light that has propagated through the living tissue, then analyze and depict functional information on the tissue. As we review the inertial sensors and PPG technology, these two main kinds of physical values are the fundamental of activity-based engagement measurement, we build the basic background and idea to how to design our services, but we still need to consider the features of users in our research. 2.1.3 The Sensors of Emotional Data in Wearable Devices. Spire, which is the first product of emotion recognition in the world, use the technology of respiration to fulfill its functions and its inventor – Moraveji introduces. 政治大. the fundamental in his paper (2011). He says the breath speed easily be influenced by. 立. different emotion, and the changes can be recorded by sensors. The most salient. ‧ 國. 學. finding in his studies is breath regulation, which reduces breath speed, activates the parasympathetic nervous system (PNS), the so-called ‘rest-and-digest’ response, to. ‧. relax the body, reducing stress and anxiety.. Philippot and Chapelle (2002) give the description about four typical emotions. sit. y. Nat. like following: . Anger: Breathe and exhale quickly; slightly deeper than regular breathing. n. al. er. Joy: Breathe and exhale slowly and deeply; breathing is very regular.. io. . Ch. i Un. v. amplitude. And breathing is slightly irregular with some tremors. . engchi. Fear: Breathe and exhale quickly; with a normal amplitude. And breathing is slightly irregular with some tremors.. . Sadness: Breathe and exhale through the nose with a normal amplitude and pace. And there are some sighs in your expiration.”. They also point that expressive emotional components like laughter or tears tend to be associated with happiness and sadness. As we review the fundamental of emotion measurement through wearable sensors, we intend to enrich the patterns of emotion that can be recognized in our research, and also consider the features of users to affirm the certain patterns.. 2.2 The Service Design in Wearable Devices. 16.

(26) We offer various services to satisfy for all kinds of consumers, but we should know the real service the customer need. Parasuraman, Zeithaml and Berry pointed in 1985, “Service quality is more difficult to evaluate, and quality is a comparison between expectations and performance.” Therefore, we should make a balance among physical, functional, and users’ features in the current technology in order to minimize the gaps between users and markets, meanwhile maximize the value of services. De Deugd et.all (2006) express the view of modeling devices as service, which is called service-oriented device architecture (SODA). SODA focus on the gaps between physical and digital realms; in the other words, a real problem can be delivered and solved through information technology such as internet with less labor cost and less space limitation. The devices range from basic sensor interfaces to complex equipment.. 政治大. The former services based on wearable device are divided into two areas: one is for. 立. non-professional needs like cellphone, another is for professional needs like ECG. ‧ 國. 學. monitoring device, whereas with the development with hardware and software, the new service between two areas is emerging. Now several enterprises have already. ‧. offered the services that focus on the fitness of daily life and are preparing to cooperation with hospitals to offer pre-diagnoses and after-diagnoses care to the. y. Nat. sit. patients of certain disease (Ghosh et.al,2013).. al. er. io. We have mentioned the physical and functional features before; we review. n. the features of baby boomers and then connect their need with wearables. According. Ch. i Un. v. to Maclnnes’ research (2006) about work-life balance of baby boomers in Europe, a. engchi. man or a woman often devotes himself or herself to work or family when they are young.. He then points that a certain number of them confuse leisure in meaning of. life after decades of bustle. According to the Health and Retirement Study (HRS), which was designed to measure assets, health, wellbeing of Americans over the age of 50, it can be found that median Boomer is richer over the past decades, yet those in the lowest quartile are less well off (Lusardi, Mitchell, 2007). They also tell us the cruel truth that is wealth distribution is quite unfair and related improvement has not been uniform. Whereas, there are still many needs not only is about material but also mental in this generation’s heart (Gilleard, Higgs, 2007). We can find some regardless truth about baby boomers that lead to the fact that there are no specialized services designed for them. We will integrate all the features, 17.

(27) and then optimize the services with the view of service design. So in our research, we not only integrate current functions from different wearable devices but also provide optimization mechanisms considering the basic functions in order to build a completed interactional service system.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 18. i Un. v.

(28) CHAPTER 3. iEnjOrange PROJECT Nowadays, people’s expectation to quality of life grows higher and higher. Health becomes a hygiene factor, a basic need. Especially for the baby boomers aged beyond 50, they have already spent most of their lives on pursuing something, and besides staying in health, “wellbeing” now is a higher goal for them to pursue. iEnjOrange plans to provide a platform that integrates services from different stakeholders and creates an ecosystem focusing not only health but also wellbeing. This chapter aims to depict the whole picture of our integrated research project－“iEnjOrange”.. 3.1 The Conceptual Framework of iEnjOrange. 政治大. In iEnjorange, we assume that wellbeing is an ultimate goal for everyone. Wellbeing. 立. is a broader concept beyond happiness (Seligman, 2011). In order to achieve. ‧ 國. 學. wellbeing, we thought there is no doubt for them to find their own meaning in life because meaning refers to a sense of comprehensibility (Park, 2014) and it is central. ‧. to wellbeing (Steger, 2012). But the word “meaning” is too abstract to realize, it needs some method to measure and calculate. Based on PERMA model, there are five. y. Nat. al. er. io. (Seligman, 2011).. sit. dimensions: positive emotion, engagement, relationship, and accomplishment. n. PERMA model is our core theory, to which we add another concept, vitality, as. Ch. i Un. v. shown in figure 3.1. It’s important to define and explain what they really work. There. engchi. are some statements of explanation as follows:. Wellbeing P. E. R. A. V. Meaning Activity. Strength. Figure 3.2. Conceptual Building Block. Note: P = Positive Emotion; E = Engagement; R = Relationship; A = 19.

(29) Accomplishment; V = Vitality. PERMAV is our core theory, as shown in figure 3.1, we need to find meaning through PERMAV. It’s important to define and explain what they really work. There are some explanations as follows: . Positive Emotion: Positive Emotion involves positive senses, such as pleasure, comfort, good,. happy, satisfied, and so on (Seligman, 2012). Positive Emotion has been considered as an important factor toward flourishing by many positive psychologists (Rogers, 1951; Maslow, 1954 & 1962; Jahoda, 1958; Erikson, 1963 & 1982; Vaillant, 1977; Deci & Ryan, 1985, Ryff & Singer, 1996; Murray, 2003; Cameron et al., 2003). . Engagement:. 政治大. Engagement is related to the sense of flow, which means one is completely. 立. self-less absorbed in activities (Csikszentmihalyi, 1997). People may feel time. ‧ 國. 學. passing quickly, and focus on what they are doing when they are in flow (Csikszentmihalyi, 1999). Relationship:. ‧. . Relationship involves one’s authentical connection with others (Seligman, 2012),. y. Nat. sit. and it is a critical factor of achieving wellbeing (Grootaert et al., 2003). Through. al. er. io. measure social capital in online and offline, it’s easier to understand people’s. n. relationship situation and achieving social wellbeing (Antheunis, Abeele and Kanters, 2015). . Accomplishment:. Ch. engchi. i Un. v. Accomplishment relates to the sense of accomplishment and successful pursuits (Seligman, 2012). According to Seligman’s PERMA model, achievement means a sense of accomplishment and success in one’s pursuits. Further, achievement is strongly linked with purpose and pursuit of important goals, which are elements of meaning of life (Park, 2014). We propose that we can reach higher degree of Accomplishment, when we well pursue meaning and purpose in life. . Vitality: In iEnjOrange, vitality means health, which represents one’s physical and. phschological status. In addtion to being the most fundamental element regarding the other four values, vitality is even more important among seniors. For example, the prevalence of reports of four chronic conditions increased significantly among 40- to 20.

(30) 59-year-olds baby boomers, including cardiovascular disease, obesity, lung problems, and diabetes (Martin et al., 2009). . Meaning Meaning involves one’s purposeful and meaningful sense (Seligman, 2012).. Meaning in life has been found a great importance in psychology, even in positive psychology, which is popular recently, leading to wellbeing. Steger in 2013 stated that the influences of Relationship and Achievement depend on character of each individual, which is affected by personal strengths. And, Achievement related to goals and purposes pursuing, which also are covered by Meaning (Park, 2014). Engagement and Positive Emotions cannot stand alone without meaning pursuing (Steger, 2013). Therefore, we put Meaning as our base of concept. . Strength. 政治大. Identifying and developing Strengths can help individuals to keep pursuing goals. 立. (Segerstom, 2001), and help them to identify meanings and purposes (Sweeney,. Activity. ‧ 國. . 學. 2014). As the result, we define strength as one of our bases under meaning.. ‧. Therapeutic recreation has developed for a long time in positive psychology. It means that specialists utilize one’s leisure experiences to enhance Strength, and reach. y. Nat. sit. the goals, which is meaning and purpose in life (Anderson and Heyne, 2012). In our. al. er. io. project, we facilitate activities intervention as leisure intervention to provide to. n. individuals and help them reach their final goals in life. Therefore, Activity is also an. Ch. important base under Meaning in our project.. engchi. i Un. v. 3.2 The Ecosystem of iEnjOrange. In order to improve wellbeing through meaning in life, it is not easy to achieve it only by a device or a service. That is why iEnjOrange project not only needs a platform but builds a whole ecosystem, which is depicted as a value network shown as Figure 3.2. The main purpose of iEnjOrange is to co-create the value and deliver them to users by the cooperation between each stakeholders and the integration of resources. As the main role between stakeholders and consumers, iEnjOrange holds massive user information, e.g. biological information and user behaviors etc. According to these information, iEnjOrange tries to find the meaning of each users and cooperates with all the stakeholders in our ecosystem to provide integrated services and 21.

(31) encourage users to achieve their life meanings. The stakeholders inside dotted lines are in the same system and directly impact the value co-creation of iEnjOrange platform. On the other hand, the stakeholders outside the dotted lines support the operation of the platform. They will be individually introduced as follows:. 政治大. 立. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. Figure 3.3. 1.. engchi. i Un. v. The iEnjOrange Ecosystem. Activity Providers In the ecosystem of iEnjOrange, activity providers are one of the most important. stakeholders, which directly or indirectly provide baby boomers with their favorite or potentially interesting activities. iEnjOrange platform can support the activities in many ways, like biological information, location information, user behavior, or communication and interaction. Users will be more willing to join or enjoy activities as activity providers can provide feedbacks to iEnjOrange, such as user behavior or status. This will result in a positive feedback cycle and gradually find the real views. 22.

(32) of users’ lives. 2.. Digital Service Providers Apps for wearable devices are the key stakeholders bridging between activity. providers and iEnjOrange. The purpose of Apps is to support users to record behaviors through wearable devices. 3.. Insurance Companies iEnjOrange plans to cooperate with insurance companies and issues new health. care plan, which focus on not only health but also wellbeing to improve the quality of life. Insurance companies provide users with smart wearable devices for free so that the users can enjoy the services in iEnjOrange platform, which are co-created by the service providers. 4.. Medical Institutions. 政治大. iEnjOrange holds massive user information so it can know the lifestyles or habits. 立. of users by analysis and professions can find meaningful insights from the analysis. ‧ 國. 學. results. This provides a way for medical and health care institutions to accurately track the patients’ life and doctors can track the changes after treatments and see. ‧. whether they are effective. This will improve medical-related institutions and their treatments in the long run.. y. Nat. Device Suppliers. sit. 5.. al. er. io. Since iEnjOrange platform highly relies on smart wearable devices, iEnjOrange. n. includes device suppliers in our ecosystem. We expect that the device suppliers can. Ch. i Un. v. cooperate with the insurance companies in leasing model so that the users. engchi. participating the new insurance plan can freely use the device and normal users can directly buy it as well. 6.. Banks & Venture Capital Companies iEnjOrange cooperates with financial institutions to encourage App developers to. develop new Apps and to lower the threshold of it. A financial institution like a bank or a venture capital company plays a role as an investor. App developers propose their ideas to the investors and those who hold potential benefits and markets can win the investme1nts from the bank or the venture capital company.. 3.3 The System Architecture of iEnjOrange. The iEnjOrange project intends to provide services to help baby boomers achieve 23.

(33) their life well-being with an integrated wearables-based platform. The platform is activity-driven and focuses not only vitality but PERMA well-being. It comprises of four components including Well-being Maximization System, Meaning System, Relationship Ameliorating System, and Positive Emotion and Engagement Optimization System (as shown in Figure 3.3). iEnjOrange Platform. Meaning Searching System. Relationship Ameliorating System. Well-being Maximization System. 政治大. 立. ‧. ‧ 國. 學. Positive Emotion and Engagement Optimization System. observation & feedback. intervention. er. io. sit. y. Nat. Figure 3.4 The iEnjOrange System Architecture. n. al. . Ch. engchi. i Un. v. Well-being Maximization System. This system integrates all the resources within the platform and generates personalized activity-based interventions through an approach that combines the results of local mechanisms to get the global optimality. It also uses an evaluation matrix to measure well-being value, a multi-dimensional concept, between each intervention. By the well-being valuation and maximization, the system tries to provide the most appropriate interventions with most expected value. . Meaning Searching System We proposed a mechanism using strength-based activity intervention to help. individuals reach their meaning value maximization. To reach the goal, we identify what users’ strengths and plans are, and adjust users’ plan through changing activities. 24.

(34) they are engaging based on their individual signature strengths. . Relationship Ameliorating System This system provides methods to improve relationship based on the concept of. social capital (Putnam, 1993). According to the behaviors on virtual and real world, system will provide solutions by the following four modules: social capital measurement, social capital increase & maximization, social wellbeing achievement and co-fulfilling verification. These modules quantify each user’s social capital and provide suggestions to maximize it. After receiving feedbacks from users and the entire platform, system will adjust to match the obtained information. In this system, we are dedicated to improve relationship and achieve social wellbeing. . Positive Emotion and Engagement Optimization System. 政治大. We use the physical data got from the wearable device to measure the positive emotion and engagement of users in a short or long term time. The real-time data can. 立. be grouped to analyze in process in order to optimize the value of data in the. ‧ 國. 學. comprehensive measurements. The ultimate goal of this system is aimed to achieve the autonomy in the circle of the whole system.. ‧. 3.4 The System Flow. sit. y. Nat. al. er. io. To maximize values which our target agents get, we provide several stages to achieve. n. the goal. There are four stages: sensing, monitoring, optimization, and autonomy.. Ch. i Un. v. In the first stage, sensing, each user has a unique preference for their well-being.. engchi. Some think relationship is the most important matter for well-being; others may think positive emotion is. Before we measure the well-being values, we need to know the preferences of each well-being value of individual user, and collect some initial input to calculate a score matrix which would be used in the second stage. To get initial input, there are some questionnaires, such as social media linked questions, device equipped questions and a training process offered by wearable device to get close view about our users. All these results can be used as references to measure PERMA scores each set of activities in the second stage we will explain in the next section. In the second stage, monitoring, our system takes one week to monitor each user daily activities and detect PERMA scores while the user is doing specific activity. Each scores detections are manage by different module: positive emotion and engagement optimizing system, relationship ameliorating system, meaning searching 25.

(35) system, and well-being maximization system. All the scores would be sent to well-being maximization system to generate a score matrix. By collecting all the activities he/she did through different content tags, the system can detect which activities might be more suitable for that user and we stored PERMA score matrix for the usage of following stages. Some constraints and limitations are necessary, such as twice date with friends a week. All modules can generate constraints and limitations based on the first stage results or PERMA concerns, and they also can be used in the next stage. In the third stage, optimization, at beginning of this stage, positive emotion and engagement optimizing system, relationship ameliorating system and meaning searching system provide some possible context tags of activities, based on the each. 政治大. system’s scores, constraints and limitations collecting from second stage. (Context tags refer to some kinds of contexts when doing an activity, which are divided from. 立. three perspectives of three systems, as shown in Table 3.1) All the context tags can be. ‧ 國. 學. integrated at the well-being maximization system, and be compared with current service of third party to calculate the recommendation checklist for the next week.. ‧. The checklist as input is sent from well-being maximization system to the other three modules, and each system take the initial input from stage one, and the scores. y. Nat. sit. collecting from second stage and the user rating are taken into account to adjust. al. er. io. average scores to customize score. Each of systems send the new score matrix and. n. new activity recommendation tags to well-being maximization system. After the. Ch. i Un. v. system receives the outputs from the other three systems, the system integrate activity. engchi. recommendation tags to get a new comprehensive checklist for that user by matching all constraints and limitations. However, the checklist of activities that we expected to create maximum value to the user sometimes is not the most suitable one. Due to the inconsistency, the system would iterate four stages to improve the suitable degree of recommendation once by once. When the user achieves maximum value, which means the whole set of activities is the most suitable recommendation for the user, the process may enter into the final stage, autonomy. Although the user achieves his/her maximum value, it is possible that the set of activities is no more top set. Due to the change of environment, interests, and feeling of freshness, all the reason might lower the value; we should go around four stages once a week.. 26.

(36) Table 3.1 Context Tags in Three Perspectives Perspective. Context Tags. Device Perspective. sleep physical exercise leisure activity sedentary activity outside activity. Strength Perspective. creativity. (Peterson & Seligman, 2006). curiosity judgment. bravery. 學. ‧ 國. 立. 治of learning 政 love 大 perspective perseverance honesty. ‧. zest. y. Nat. love. social intelligence. n. al. Ch. i Un. teamwork. e n gfairness chi. er. io. sit. kindness. v. leadership forgiveness humility prudence self-regulation appreciation of beauty and excellence gratitude hope humor spirituality. Social Perspective. independent. 27.

(37) double grouped crowded private public virtual realistic outdoor indoor. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 28. i Un. v.

(38) CHAPTER 4. THE ACTIVITY-BASED ENGAGEMENT OF BABYBOOMERS MECHANISM THROUGH WEARABLE DEVICES 4.1 Conceptual Framework This section provides the conceptual framework of our research. There are six significant concepts in the conceptual framework: service credibility, wearables usability, interaction frequency, involvement degree, user data value, optimization level and its arrow relationship (see Figure 4.1).. 立. 政治大. ‧. ‧ 國. 學. Figure 4.1 Conceptual Framework Diagram. sit. y. Nat. . Sensor-based Data Service: It refers to ease of use, main functions and. io. n. al. er. appearance designing of our wearables, which can enable to realize our service. i Un. v. from the Internet and send feedback information to the internet. Wearables which. Ch. engchi. have some sensors are the media that we connect with users in real world, which is an another significant element of whole interaction process. If the value of sensor-based data service is high, the user will wear the device more times in order to enrich the data that concerns to the user’s condition. . Activity-based Engagement Service: It refers to the connection between physical data and life style, which can help understand comprehensive life about oneself. It is the information media that we can build the connection with users. If service in our system is high, a user will give more accurate details at any stages that contribute to an accurate measurement and an appropriate suggestion we can offer.. . User data richness: It refers to two aspects of data; on the one hand, a same pattern of data in various time spaces in order to get the changes and trend; on. 29.

(39) the other hand, various pattern of data at the same time in order to improve the total measurement accuracy. . User Information Accuracy: It refers to the final information we measure through data when users wear their device, which determines the first analysis effect. Only the level of accuracy is high, can we describe a user correctly using data.. . User Information real-time: It refers to the frequency users wear their device and update their data into internet, which has a close relationship with effective quantity of information exchanges between users and our system. If data can be update in time, various data are recorded that will help to analyzes users comprehensively.. . Optimization level: It refers to a comprehensive measurement that covers both. 政治大. mental and physical conditions through the process and analysis of sensor data. In our research, it mainly aims at positive emotion and engagement. . 立. Intervention effect: It refers to the way the system give suggestion to users that. ‧ 國. 學. mainly considers the timing and the content. The intervention, which plays an important role in the process of autonomy, is the direct connection between users. y. Nat. Arrow (1) Sensor-based data service influence on user data richness: Due to. sit. . ‧. and the system, so that its effect represents effect of the whole system.. al. er. io. battery using time, there is a three-way trade-off between performance, cost and. n. power consumption. The dilemma leads to the condition that we hardly record. Ch. i Un. v. the whole data in real time. Fortunately, increased sensors in the wearable. engchi. devices lead we can read various data at one time that means we can measure more accurate. . Arrow (2) Sensor-based data service and activity-based engagement service’s influence on user information accuracy: Physical data become meaningful information for the user to understand their life and what they can do to live with happiness.. . Arrow (3) Sensor-based data service and activity-based engagement service’s influence on user information real-time: We believe that these services make a road that the data can run on it quickly and timely from user to our system.. . Arrow (4) User data richness’s influence on user information accuracy: The quantity of data we can collect is limited in the constant time, so we need high-quality data. Then it is possible to analyze the high-value information. 30.

(40) . Arrow (5) User data richness’s, user information accuracy’s and user information real-time’s influence on Optimization level: Optimization here is an algorithm, which is about measuring positive emotion and engagement and integrating various patterns of data to get a comprehensive result. The technology of data mining will be used to make a classification decision in this process. The much richer, more correct and more timely data are, the more possibilities to identify relevant objects of user.. . Arrow (6) Optimization level’s influence on Intervention effect: The service, which is based on prior data from sensors, is tailored uniquely for each user, and it is traditionally made by manual work. However, automated intervention is largely relied on former precise result, that means only optimization level is high. 政治大. can we offer a perfect intervention effect. Here, we argue that optimization level will positively influence intervention effect.. 立. . ‧ 國. 學. 4.2 Design logic and research approach. The system architecture that is connected with the conceptual framework is shown in. ‧. the section. The system architecture provides the design and implementation with the details that are the specific method to fulfill the conceptual framework provided in the. y. Nat. sit. previous section 4.1.. al. er. io. The system architecture consists of four modules and four databases, which are. n. activity-based sensing module, activity-based monitoring module, activity-based. Ch. i Un. v. optimization module, activity- based autonomy module, sensor database, user basic. engchi. information database, time database and preset value database (see Figure 4.2). The activity-based sensing module can be viewed as a mapping to the concept of service credibility and wearable usability described in the conceptual framework. The activity-based monitoring module can be considered as the fulfillment of interaction frequency and involvement degree. The activity-based optimization module, which measures positive emotion and engagement depending on users’ data and coming from the concept of user data richness and optimization level. At last, activity-based autonomy module defines intervention effect from the system to users and feedback to the system.. 31.

(41) Figure 4.2 System Architecture These modules are briefed below and will be detailed in the following. 政治大. subsections:. 立. (a) Activity-based Sensing Module: The complete user database, which consists of. ‧ 國. 學. basis information and sensor data, are classified and stored in our system. As a result, basic functional data (ex. heart rate, breath rate, posture) can be attained at. ‧. this module, and unique user label will be connected with specific user. (b) Activity-based Monitoring Module: According to previous information, seven. sit. y. Nat. main functions, which are nutrition, sleep, sport, health, calories, emotion and. io. er. activity, will be attained at this module. Activity, which concerns to recognize the activity pattern, is needed to detailed activity information from the help of outside. n. al. Ch. i Un. v. social resource. Moreover, emotion, which concerns to recognize emotion, is an. engchi. innovative measurement compared with physical health in order to achieve the goal of comprehensive health. (c) Activity-based Optimization Module: It refers to the algorithm that calculates scores of positive emotion and engagement. Then activity patterns as the prediction will be given from the sores and they also serve as the feedback that will be returned to the system, yet users will not get the result at this phase. (d) Activity-based Autonomy Module: More than one user’s information will be integrated in order to give more valuable and more specific suggestions to every certain user. Suggestions, whose contents are gotten from real-time analyzed result in wearables, are needed to confirm the time and way to intervene, when we consider the differences among users. Moreover, it is a connector to the whole interactive system, that means, information of users, which is about effect and. 32.

(42) improvement, will be exchanged here.. 4.3 Activity-based Sensing Module The original data from sensors, which are continuous analog signals, must be sampled by several rules in order to convert into digital signals. With the process of data organization and data mining, digital signals finally become useful data (see as Figure 4.3). Available sensors can capture diversity of data, like inertial measurements and heart rate from the human body. The types of data must be uniquely tailored for each certain function that can be performed by one or several sensors. Here, heart rate, breath rate, location, time, and posture detection as raw data (see details as Table 4.1),. 政治大. which sync with users’ smartphone through Bluetooth, are transmitted to the system. 立. through networks.. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. Figure 4.3 Body Sensor Networks (Sazonov et.al, 2014). 33.

(43) Table 4.1 The Relationship between Basic Functions and Sensors BASIC. SENSOR. MECHANISM. FUNCTION HEART RATE. LED. Assess the arterial pulsatility skin vascular beds. BREATH. Inertial. Detect the respiratory movement of body:. RATE. sensors. expansion. and. contraction. of. your. abdominal and/or thoracic cavity LOCATION. GPS. Confirm the location with the help of electronic map. TIME. Bluetooth. POSTURE. Inertial. DETECTION. sensors. 立. 治政 Sync with the smartphone 大. Detect body motion acceleration range &. ‧ 國. 學. angular velocity & body movement. ‧. In iEnjOrange, we use “Fitbit Surge” (www.fitbit.com) and “Spire” (www.spire.io), which can sense all the required data, instead of designing our own. sit. y. Nat. wearables. Why we choose Fitbit Surge and Spire is based on several reasons below:. io. er. (a) Fitbit Surge, which is a professional fitness tracking wearable devices, can not only track all-day stats of activities and sleep but also record continuous, automatic,. n. al. Ch. i Un. v. wrist-based heart rate. Moreover, its large display is great for check real time stats like. engchi. distance, time and pace. The Fitbit Company says, “The sleek-looking Surge is Fitbit’s most advanced tracker to date.” (b) Spire, which is a new kind of tracker differing from traditional wearables, can track users’ emotion through respiration. It tracks and improves users’ state of mind by allowing them to discover when they are stressed, where it happens, and what they were doing. It is a try to measure both physical and mental health. Fitbit is worn as a watch on the wrist (see Figure 4.4 (a), (b)), while Spire is worn closely to skin of waist or breast (see Figure 4.5 (a), (b)). 34.

(44) . Figure 4.4 (a) Fitbit Surge’s Appearance Figure 4.4 (b) Fitbit’s Wearing Style. 立. 政治大. ‧. ‧ 國. 學 sit. y. Nat. n. al. er. io. Figure 4.5 (a) Spire’s Appearance Figure 4.5(b) Spire’s Wearing Style. Ch. 4.4 Activity-based Monitor Module. engchi. i Un. v. This module is designed for recording the daily life of the user, which mainly includes activities and emotions, and for organizing data of former sensing module by every activity.. 4.4.1 Activity When talking about various activities users do every day, we can count thousands of activity names so that the big challenge in this part is making a balance between activity patterns and sensor technologies. The final activity patterns we can recognize (shown as Figure4.6) are eight activities that are Sleep/Wake, Work/Study, Physical exercise, Entertainment, Home stay, Travel, Hobbies and Constant time activities. We 35.

(45) will introduce the heuristic of recognizing every activity pattern in details shown below.. 立. 政治大. ‧. ‧ 國. (a) Sleep/Wake. 學. Figure 4.6 Recognized Activity Patterns. This activity can be easily recognized by heart rate. When a person is in the sleep. y. Nat. condition, his or her heart rate will be lowest stage of the whole day as the red line in. n. al. er. io. sit. the figure 4.6 (the sleep time is 2:34-8:28 and 14:50-17:13 on that day) shows.. Ch. engchi. i Un. v. Figure 4.7 The Heart Rate Changing of One Person in The Whole Day. The quality of sleeping, which is another function most users care about, can be recognized through changing times of postures as figure 4.7 shows. Figure 4.7 (a) has. 36.

(46) more times of changing postures than Figure 4.7 (b) that can be concluded as “The sleep quality of Figure 4.7 (a) is worse than another.”. (a). (b). Figure 4.8 The Posture Changes in The Sleep Condition. 立. (b) Physical exercise. 政治大. We define the physical exercise as following （reference the function of heart rate and. ‧ 國. 學. steps from Fitbit）:. We encourage the user to take exercise at scattered time to reach the goal of. ‧. 10,000 steps per day.. sit. y. Nat. (c) Leisure activity. io. al. n. steps from Fitbit）:. er. We define the physical exercise as following （reference the function of heart rate and. Ch. i Un. v. The heart rate is between 70 and 100, meanwhile steps must more than 100 per. engchi. hour in order to make sure you are not always sitting on the chair. (d) Sedentary activity We define the physical exercise as following （reference the function of heart rate and steps from Fitbit）: The heart rate is between 50 and 70 except the sleeping time, meanwhile steps less than 100 per hour. (e) Outside Activity We encourage the users have more time to go out to enjoy the fresh air and close to the nature.. There is still a training process that will record customs of the user (Specific. 37.

(47) activity with time and location) before the real activity recognition, and it leads to a personalized activity schedule, which will help to analyze data in following process. In the training process, more interaction among the user, the wearable, and the system will exist.. 4.4.2 Emotion According to previous research, there is a clear relationship between respiration and emotion as Figure 4.11 shows. The basic mechanism of Spire is the breath rate will be influenced by activity and emotion (Moraveji et al, 2011) and Spire has correctly recognized three emotion patterns -clam, tense and focus.. Philippot and Chapelle. (2002) had pointed that expressive emotional components like laughter or tears tend. 政治大 infer we can measure the emotion 立 of happiness and sadness in a certain time through. to be associated with happiness and sadness. So based on the same mechanism, we. ‧ 國. 學. the times of laughing and crying. The specific emotion pattern, which is recorded with time and location, will be compared with the specific activity pattern in the latter analysis.. ‧. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v. Figure 4.9 The Relationship between Respiration and Emotion. The Spire we use mainly measure which state the user is in according to the. 38.