The data will be collected using the SenseWear Pro continuous body monitoring device made by BodyMedia.
The BodyMedia software contains the algorithms to detect different levels of activity. There are many variables that we can use to judge if a person has been exercising or not. These include:
- Number of steps taken
- Energy expenditure (calories)
- Average Metabolic Equivalent of the Task (kcal/hr/kg)
- Active energy expenditure (3 METS or higher)
A typical summary of the data looks like this:
Data can be graphed for more in-depth analysis:
(Explanation of data)
- A - Walking to get lunch
- B - Walking to car
- C - Walking to class from parking spot at CMU
- D - Walking back to car
- E - Running for 20 minutes
Data can be exported to three different types of files:
We will collect the user's data for a whole week prior to giving them the robot. Therefore we will have a baseline of their typical performance. This will allow us to set their goal for the next week.
One challenge is that when the data is uploaded, the user will be able to see his or her data numbers when they are exporting the file. A solution to this would be to automate the data upload, or to get a version of the software that exports to a file without revealing the analysis. This file can then be read by a program that controls Sporto's actions.