Press 'Tab' to the content


24 MAR 2016 Seminar

Series of Dept Research Seminars - "Development and Evaluation of a Pre-impact Fall Detection, Classification, and Protection System with Wearable Airbags for Elderly" (Date: 29 March 2016)

Probationary presentation by Miss Kaifeng Liu

Probationary presentation by Miss Kaifeng Liu

IMSE PhD candidate, Miss Kaifeng Liu will give her probationary presentation. The seminar information is as follows. All are welcome to the seminar.


Background: The rapid aging of the world population lead to a rise of proportion of elderly individuals. Meanwhile, the proportion of elderly who living alone or only with their spouse have been increased, which pose great challenges to the healthcare of elderlies. Fall accidents are the most commonly encountered problem faced by elderly people and result in both physical injuries and psychological illnesses. In order to reduce and prevent the injuries in the elderly form falls, fall detectors and their hardware are largely studied in recent years.

Objective: There are three main objectives in the present proposed study. First, the study aims at developing a wearable airbags system for pre-impact fall detection, classification, and protection for elderly. The main functions of the system are to: i) detect fall events in the descending phase; ii) classify the fall types (e.g., fall direction and fall speed) and make the judgment of the probable parts of body that may hit the ground or get hurt during the fall; iii) provide cushion protections by inflating the airbags at corresponding positions of the body. Second, a systematic human factors approach will be developed as a common evaluation framework for all of the fall detection and protection systems. The developed system will be evaluated from both the technology and users’ perspective. Third, the study will examine factors affecting the acceptance of the developed system. 

Method: For the development of the system, a three-axis accelerometer and a gyroscope will be attached to the waist of the participants and used to collect movement data of different types of falls and activities of daily living (ADL). The algorism will be designed with machine learning method (MLM) to distinguish fall events from ADLs and classify different kinds of falls as well. The hardware, which includes the airbags, battery, gas cartridge, trigger mechanism to release the gas, inflator to inflate the gas will be integrated to the system along with the sensors. For the evaluation of the system, the whole system will be tested by recruiting participants to perform different types of scenarios and key performance indicators (KPIs) such as sensitivity and specificity will be used to describe the performance of the system. A systematic human factors evaluation method will be designed and employed to further identify the usability issues of the system. Furthermore, in order to better understand the elderlies’ acceptance of the technology, end user interviews will be performed to figure out the facilitators and barriers that would influence the adoption of the technology.

Discussion: This study could provide a fresh perspective that the classification of falls should be involved in the fall detection and protection studies to make the system better understand the fall direction and speed, and provide more effective and targeted protections. In addition to the cushion propose, the rescue will also be more effective if the system could deliver more meaningful and detailed information of the fall event to the caregivers or emergency contacts. The design of the system will come one step closer to improving the prevention of the fall and the emergency responsiveness as well for fall related trauma. Moreover, the study will provide a systematic human factors evaluation method, which also take users perceptions and acceptance into consideration. The evaluation method could be used as a common framework to evaluate all kinds of fall detection and protection systems.  


29 March 2016 (Tuesday)


15:30 - 16:30




Miss Kaifeng Liu, IMSE PhD candidate


Dr. Calvin K.L. Or

< Back