Collect or provide high accuracy data only when strictly necessary.
When collecting or displaying data, it is tempting to use high resolutions. The problem of using data with high resolution is that its collection and manipulation require more resources (e.g., memory, processing capacity, etc.). As a consequence, energy consumption increases unnecessarily.
For every use case, find the optimal resolution value that is required to provide the intended user experience.
Consider a running app that is able to record running sessions. While the user is running, the app presents the current overall distance in real-time. While calculating the most accurate value of the total distance would provide the correct information, it would require precise real-time processing of GPS or accelerometer sensors, which can be energy greedy. Instead, a lightweight method could be used to estimate this information with lower but reasonable accuracy. In the end of the session, the accurate results would still processed, but without real-time constraints.