Senior Design Projects
ECS193 A/B Winter & Spring 2020
Citizen Science: Mobile Device Application for Collection of Human Gait and Mobility Data
| ********** | |
| Erik Henricson. PhD and Lisa Miller. PhD | |
| UC Davis Physical Medicine & Rehabilitation & Human Development & Familty Studies / Adult Development Lab |
Project's details
| Citizen Science: Mobile Device Application for Collection of Human Gait and Mobility Data | |
| In typically-developing people. stride length and step cadence increase with increasing level of exercise. Patterns of walking activity change as we age or if we are affected by diseases that impair our mobility. People show impairment in walking ability and efficiency due to events such as strokes. due to muscle weakness from diseases (such as muscular dystrophy). or even due to typical loss of muscle with advanced age. Many of the tools that are used to measure these impairments are clinical tests administered by health care professionals. and some tests require specialized and expensive equipment. At the same time. in the form of mobile smart phone devices. the general population has daily access to a suite of highly accurate and technologically advanced motion and location sensors that have been largely overlooked by the medical research community. We have demonstrated that accelerometer and GPS data from mobile phones and other small sensors can describe patterns of walking motion. speed. and distance and range of travel that could be useful in evaluating and monitoring people’s walking ability in the home and in the community across a wide range of disorders where a person’s mobility is limited. The figures below (see attached PDF file) represent data collected using an iPhone and multiple apps running simultaneously. Figure 1 shows patterns of total 3-d XYZ axis acceleration (G’s) from a phone accelerometer during typical walking. Figure 2 shows the visibly different patterns of lateral accelerations from the same Figure 1 data (top panel) and from an impaired walker (bottom panel). Figure 3 shows samples of accelerometer data representing different walking speeds. paired with simultaneously collected GPS positional data showing speed. distance and range. | |
| In order to “train” machine learning-based systems to detect atypical walking patterns using single accelerometer sensors. we require a large volume of representative “typical” walking data from the general population. We seek to develop a prototype for a “citizen science” mobile device application to allow volunteers to record data on typical walking as measured by mobile device accelerometers and to transmit that data to a central data repository. | |
| For this project. we challenge a team to develop a straightforward mobile device application (on iOS or Android Platforms) to collect and log 3-D g-force data from a mobile device’s accelerometer (at 50-100Hz) as a time-stamped .CSV file. along with time-stamped GPS data in the form of a .GPX file. The application should be capable of labeling the files with a person’s unique study ID number. and also of forwarding these files from the device via email or via a “cloud” server function. The ability to attach a short video to the files to show a person’s walking pattern would be desirable but is not required. This is part of an ongoing program of study at UC Davis. Scientific/academic acknowledgement and intellectual property rights for students will be governed in accordance with University policy. |
|
| 1. Familiarity with mobile device application developer’s environments (iOS or Android). including dynamic access to device internal sensor data and GPS positioning data 2. Familiarity with principles of cloud data transmission from the mobile device environment |
|
| ********** | |
| 30-60 min weekly or more | |
| Client wishes to keep IP of the project | |
| Attachment | Click here |
| No | |
| Team members | N/A |
| N/A | |
| N/A |