2018 Fiscal Year Research-status Report
Wear-I: A Multi-Wearable Organic System for Smarter Individual Services
| Project/Area Number |
18K11408
|
|---|
| Research Institution | Hosei University |
|---|
Principal Investigator |
Jianhua Ma 法政大学, 情報科学部, 教授 (70295426)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
Huang Runhe 法政大学, 情報科学部, 教授 (00254102)
|
|---|
| Project Period (FY) |
2018-04-01 – 2021-03-31
|
| Keywords | Wearable / Platform / Modeling |
|---|
| Outline of Annual Research Achievements |
In 2018 fiscal year, our main research was focused on development of a multi-wearable platform and representative applications in using various wearable devices.
A fog-cloud integrated framework has been adopted to develop the multi-wearable platform manage diverse devices, their states and their data. The platform uses DynamoDB of Amazon Web Service (AWS) in the cloud for universal data management and local server as fog to manage data from various devices smart watch and myo for fast response. We also conducted a preliminary study to use the blockchain technique to protect data privacy.
We have studied the data quality from multiple wearable devices and possible impact to human activity recognition with considering devices’ misplacement and temporal differences. Multi wearables have also been exploited for users’ behavioral modeling such as preference, and human psychological state analyses such as emotion features and personality computing.
These research results have been published in two internal journals and presented in three IEEE international conferences, respectively.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
The research was smoothly carried out as we planned. Since our laboratory had already possessed many wearables, the design and development on the multi-platform were able to start in the beginning. Although data from wearables are kept permanently in the remote cloud of Amazon Web Service (AWS) with DyanmoDB, we also have used a Linux CentOS based server inside our laboratory. The server is functioned as a fog with MongoDB for quicker response to nearby wearables. A series experiments were conducted to test the performance, speed and scalability using various wearables and sensors.
One basic characteristic of this research is to use multi devices for many different applications, which were fallen into two categories. One is to combine smartphone, smart watch and Myo to achieve higher recognition rate of physical activity such as walking and sitting. The other is to use biological wearables including Muse and Spire to recognize emotional states. When using multi wearables, it is necessary to synchronize multi data streams from different devices. Therefore, we made many experiments to study the data quality including temporal differences and their impact to activity recognition.
|
| Strategy for Future Research Activity |
The research in FY 2019 will be carried out from the following three aspects.
A data security scheme will be added to the multi-wearable platform. We are going to use the blockchain technology to control data access and update so as to achieve data operation traceability. All information about users and their multi wearables devices will be protect as well.
The approaches and techniques to effectively use multi devices and their data will be one of important research focuses in 2019. A relatively simple or atomic activities that will be recognized using a single wearable, and a complex or high-level activity can be further recognized by combining and fusing a set of the simple activities. A proper hierarchical mode with hybrid recognition algorithms will be explored and tested.
Human psychologic state recognition and personal character computing will be studied with using various wearable sensors including inertia, physiology and brainwave. The combinational uses of these wearables is for higher recognition accuracy by data fusion and reasoning.
|
| Causes of Carryover |
One main reason is that we just used once for traveling to present our research achievement in an international conference. Personnel expenditure and miscellaneous expenses were used less as planned.
In year 2019, we will be able to have more research output, which will be published and presented. So, it is expected to spend more budget in publishing our papers in journals and present our research in attending conferences. We shall have more research exchange with other research by mutual visit and discussion.
|
