Human Movement Mapping

Research from a wide range of studies indicate that engaging in regular physical activity (PA) has positive effects on health outcomes. Travel, particularly commuting, consumes a significant proportion of the daily activity in which people engage. Driving, in general, is the least physically demanding travel mode and contributes to modern sedentary lifestyles and the associated health problems. Using public transport, as well as walking and cycling, provides opportunities for individuals to increase their levels of PA on a regular basis. The goal of this study was to test and evaluate the role of wearables (smartwatches) in combination with data collection instruments such as activity diary to gain an in-depth understanding of travel-related PA. Smartwatches provide an accurate spatial and temporal account of daily activities (GPS tracks of the movements), as well as supplementary contextual data (e.g. heart rate, pace/speed, calories, temperature) not commonly collected in travel studies. Additional benefits refer to the use of devices that passively collect data, thus reducing the respondent burden, while shifting the load to the analysts who integrate and calibrate these data sources. Yet, the participant compliance is key, and most of the smartwatches require the user to activate a widget prior to undertaking the activity. There are also situations when the battery or indoor activities prevent collecting full data 24/7. This means that in most situation, some data is missing or is incomplete, and the analyst needs to integrate the data from other sources, such the diary. Currently, there is data from participants in office-based (sedentary) occupations available to analyse and model, but they include only two days of activities, with limited details, for confidentiality reasons. Students are encouraged to collect their own data for a week of activities and combine it with a light activity diary, to reconstruct data about all their activities, which enables easier cross-validation of GPS and PA data with the diary reports. The analysis will offer insights on the contribution of travel to PA, but also a better understanding of activity scheduling and travel mode choices, increasingly more relevant in the COVID-19 circumstances.

The data gathered for this project will be collected by the students themselves and not used for research purposes. The idea is to develop software tools that could be deployed for a future project with appropriate ethics approvals. Note this project is not to develop a fitness tracker app (like Strava), but rather a way of recording a limited set of types of activities undertaken, their locations and time spent, fusing the information from Garmin/Apple with the light activity diaries.

The software will (priority order, more important functions first):

1. Record daily GPS trace of personal physical activities and movements using a suitable wearable device (smart phone or fitness wearable).
2. Label significant places (e.g., work, home, school, shopping centre, entertainment, etc.) and/or types of activities for these traces.
3. Label modes of transport between locations (e.g., car, train, bike).
4. Support integration with a 'light' user diary of user labels for the activities performed.
5. Identify and log any anomalies in the data or missing values.
6. Integrate the above streams into a dataset that can be output for research use.
7. Display a labelled map of activities.
8. Display summary information about activities (e.g., time spend walking, calories used during activity).

Client

IP Exploitation Model

Department of Computer Science & Software Engineering The University of Western Australia Last modified: 27 July 2021 Modified By: Michael Wise