Energy consumption

Code

from libs.chapter4.analysis.battery import mean_consumption_per_device
from libs.chapter4.analysis.data_loading import load_battery_results


battery_df = load_battery_results()

Analysis

Table 11.1 shows the estimated energy consumption (% and mA) for each configuration and device (i.e., C1 and C2). The average consumption per TUG execution using the system with the C1 configuration is approximately \(0.01\%\) and \(0.005\%\) of the total battery of the smartwatch and the smartphone respectively, which is around \(0.058mA\) and \(0.254mA\), yielding a combined consumption of \(0.312mA\). In C2 configuration, the system consumes \(0.006\%\) of the smartphone’s battery, which equals \(0.307mA\). While both configurations report a similar consumption, C1 is limited by the consumption of the smartwatch device with respect to C2 (i.e., the smartwatch’s battery would run out before the smartphone’s).

Code

mean_df = mean_consumption_per_device(battery_df)
mean_df.round(3)

Table 11.1: System’s energy consumption per TUG execution on its two configurations.

		consumption (%)	consumption (mA)
configuration	device
C1	sw	0.010	0.058
C1	sp-paired	0.005	0.254
C2	sp	0.006	0.307

Summary

These reported consumptions would allow to hypothetically run (without taking into account the consumption of other services running in the devices) thousands of TUG executions with a single battery load: \(+10,000\) in C1 and \(+16,000\) in C2. Therefore, we consider that the consumption of the developed system is low, and in both cases the need for performing sufficient TUG tests on a single battery charge is comfortably covered.

Tip

The documentation of the Python functions employed in this section can be found in Chapter 4 reference: