Measurement of Activity Patterns among Wheelchair Users via GPS and Wheel Rotation Logging Devices

Dan Ding, PhD, Rory A. Cooper, PhD
Department of Rehabilitation Science and Technology, University of Pittsburgh
5044 Forbes Tower, Pittsburgh, PA 15260
Human Engineering Research Laboratories, VA Pittsburgh Healthcare System

ABSTRACT

People with disabilities tend to lead sedentary lifestyles than their unimpaired peers. Information describing physical activity profiles of wheelchair users is very limited, especially in real-life settings through instrumentation. The study combined a GPS (Global Positioning System) device and a wheel rotation logging device to measure activity patterns among wheelchair users. Bench trials showed that activity scope, intensity, frequency, and duration can be quantified and outdoor travel and use of transportation can be derived. The instrumentation provides an objective means of capturing activity patterns of wheelchair users, and enables further investigation on device effectiveness, exercise compliance, early problem detection and intervention.

KEYWORDS

Wheelchair, GPS, wheelchair rotation, physical activity

INTRODUCTION

Results from population surveys indicate that people with disabilities tend to lead more sedentary lifestyles compared to their unimpaired peers [1]. According to Healthy People 2010, based on data from the 1997 National Health Interview Survey, as much as 56% of adults with disabilities did not engage in leisure time physical activities [2]. The lack of physical activity can result in physical deterioration and impairment of multiple physiological systems such as reduced cardio respiratory fitness and impaired circulation to lower extremities [3]. To date, there is little information describing physical activity profiles of individuals with disabilities, thus limiting supportable health improvement guidelines. While many methods are available for assessing physical activity in able-bodied people, they have limited utility for people with disabilities, especially for wheelchair users. 

Similar to a pedometer that measures walking distance, we have developed a wheel rotation logging device to record time stamps of wheel rotations, and provide activity summaries of wheelchair users, e.g., traveling distance, velocity, duration, and number of stops during any period of data collection [4]. The device has been used to collect activity data of wheelchair users in free-living environments, and one of the studies [5] reported that 39 manual wheelchair users traveled 1994 ± 1851 meters and were active for 7.1 ± 4.9 hours per day in their home environments. In this study, we integrated a Global Positioning System (GPS) device with the wheel rotation logging device with the purpose of extracting detailed activity profiles of wheelchair users in terms of scope, frequency, intensity, and duration. Real life information through instrumentation eliminates the possibility of recall bias and misinterpretation of survey questions [6] , and provides quantifiable measures for investigating device effectiveness, evaluating exercise compliance, early problem detection and timely intervention.

METHODS

An off-the-shelf GPS board (u-blox Ltd.) was used to develop the GPS logging device which collects traveling information of wheelchair users including longitude, latitude, altitude, speed, and time and stores them in the flash memory (see Figure 1) . As power consumption is of primary concern in this study for a sustained period of data collection, the GPS logging device was configured to the TricklePower mode which enable s a position fix at any configured interval. Time and distance filters can also be programmed to enable a position fix at longer intervals if the wheelchair is not moving.

The GPS device is able to calculate its position by picking up radio signals from various satellites orbiting earth. However, if the "line of sight" from the receiver to the satellites is blocked, the signal degrades and may become unusable. Especially the GPS device will become useless in some indoor environments as it cannot retrieve a satellite signal. In order to compensate for GPS data loss and extract detailed activity profiles of wheelchair users in both indoor and outdoor environments, we incorporated data from the wheel rotation logging device into the analysis. Four aspects of the activity profile are to be examined including scope, frequency, intensity, and duration. The activity scope is defined by the distance from home to the furthest place. Given the latitudes and longitudes of the known locations of home (lat1, lon1) and the furthest place (lat2, lon2), the distance can be calculated as follows,

Distance (meter) = d * 180/3.14 * 60/ 1.15077945 * 1600

In terms of frequency, number of trips away from home can be extracted from the GPS data. Activity intensity defined by the traveling distances and speeds in indoor and outdoor environments respectively can be obtained by analyzing data from both the GPS and wheel rotation devices. Similarly, activity duration in terms of time spent on driving wheelchairs in both indoor and outdoor environments can be calculated. Transition from outdoor to indoor environments is identified by a certain amount of time elapsing between GPS position fix, e.g., 5 minutes. The custom program for the wheel rotation device then takes the time stamps of these position fixes and calculates distance, speed, and duration traveled indoors. If there is no activity recorded by the wheel rotation device between such position fix, the duration may indicate the use of motor vehicle transportation by wheelchair users either dependently or independently. In the case that GPS device can pick up signals inside a vehicle, the speed and distance traveled between position fix are examined to distinguish vehicle travel and wheelchair travel.

RESULTS

Photo 2: The GPS coordinate map showing the activity (Click image for larger view)

Several bench trials were conducted using a manual wheelchair with both GPS and the wheel rotation logging device (see Figure 1). In one trial, the investigator tested the system for about two hours. Figure 2 shows the map where all the travel routes recorded by GPS were marked. A custom MATLAB program was used to analyze the data from GPS and the wheel rotation logging device. The activity scope was calculated as 1527 meters from the start (usually home). Based on the assumption that large time intervals occur during the outdoor/indoor transition (i.e., 5-minute was used in the program), there were 2 trips away from home. The time stamps of the two trips were fed into the MATLAB program to calculate the total distance, average speed, and duration traveled indoor versus outdoor based on wheel rotation information from the logging device (see Figure 3). Figure 4 shows the speed between GPS position fix, and the accumulated distances from the wheel rotation logging device. There were two periods of speed data over 4 meter/second and no distance was traveled during these periods, indicating the use of motor vehicle transportation.

DISCUSSION

Graph 1: Indoor versus outdoor activity summary (Click image for larger view)

Unlike a pedometer that detects only total steps over the observational period, the instrumentation developed in the study provides an objective means of capturing activity patterns of wheelchair users. The examination of activity patterns among wheelchair users is critical for determining current levels of activity, understanding the dose-response relationship between activity and health, and determining the effectiveness of intervention programs designed to improve physical activity. The frequency of trips and outdoor activities may also indicate their community participation. More activity variables, such as duration at low, moderate, and high intensity, and day-to-day variability and weekly rhythms in activity patterns could be included in later studies to quantify activity profiles in greater details.

Future work will also focus on refining and further validating the analysis program with longer trials (e.g., 3-6 days), and correlate the results with daily dairies. Battery selection, power consumption, and package size of the GPS device need to be balanced to enable sustained trials and easier attachment to wheelchairs as well. The system will be eventually tested on wheelchair users to quantify their daily activities.

REFERENCES

1. Brown DR, Yore MM, Ham SA, Macera CA. Physical activity among adults >or=50 yr with and without disabilities, BRFSS 2001. Med Sci Sports Exerc . 2005; 37(4):620-9.
2. US Department of Health and Human Services. Healthy People 2010: understanding and improving health. 2 nd ed. Washington, DC: US Government Printing Office, 2000 Nov.
3. Durstine JL, Painter P, Franklin BA, Morgan D, Pitetti KH, Roberts SO. Physical activity for the chronically ill and disabled. Sports Med 2000 ; 30(3):207-19.
4. Spaeth DM, Cooper RA, Albright S, Ammer W, Puhlman J, Development of a miniature datalogger for collecting outcome measures for wheeled mobility, 2004 RESNA Annual Conference Proceedings , CD-ROM.
5. Tolerico M, Investigation of mobility characteristics and activity levels of manual wheelchair users in two real world environments, MS Thesis, University of Pittsburgh, 2005.
6. Tudor-Locke, C.E., and Myers, A.M. (2001), Challenges and opportunities for measuring physical activity in sedentary adults. Sport Medicine , 31(2):91-100.

ACKNOWLEGEMENTS

The work is supported by NIDRR H133F040006, PVA 2264-01, and VA Center of Excellence for Wheelchairs and Associated Rehabilitation Engineering B3142C.

Author Contact Information:

Dan Ding
Human Engineering Research Laboratories
VA Pittsburgh Healthcare System 151R-1
7180 Highland Drive, Pittsburgh, PA 15206.
412-365-4850
dad5@pitt.edu