Purposefully Optimising for Battery Life

A basic experiment in how far sustainable behaviours can take us when using modern mobile devices.

Based on the last piece on battery life I decided I would run an experiment to see how long I could actually run my phone for.

Accidentally Optimising for Battery Life

In the interest of doing somewhat of a science, I ended up taking measurements and tried to make this repeatable for others.

Hypothisis

Through changing user behaviour such as minimising screen time and network usage, android devices should have a battery life greater than the manufactures specifications, even with some level of battery degradation.

Preperation

The Device

Make/Model: Moto Z2 Play

Manufacturer specified battery duration: Up to 30 hours

OS: Lineage

GApps: None

Settings (for duration of experiment unless otherwise stated):

Brightness: 0%

Battery saver: On

Bluetooth: Off

Location: Off

Wifi: Off when out and about

Mobile network: On

Airplane mode: On during the night

Auto-rotate: Off

All other settings: Default

The methodology

I did not inform anyone else that I was running this experiment so that any calls or messages I recieved were not artificially reduced.

I allowed myself to use any other distractions during this experiment, including using other devices such as a laptop to minimaize time using the phone.

The phone was charged to 100% overnight and taken off charge at 1000. The experiment lasted till the battery level reached a point I was uncomfortable continuing without charging.

Airplane mode was enabled over night at around 2200 and turned off around 1000. All other settings remained the same.

The Data

An average day

The experiment

At about 90%, the the phone was estimating about 3 days left. This could be the possible maximum with the current battery health. At this point the screen had not been on much and I had recieved no calls, so to reach this maximum the phone would have to essentially be an internet enabled paper weight for the full 3 days.

On the second morning the phone was at 51%, with an estimated 1 day 8 hours left. This day ended up being a real struggle, with me needing to use a few apps to message people and read documents. In hindsight I possibly could have sent the links to my laptop and read them there. 106 minutes worth of voice calls ended up causing the free fall in the graph. This was mainly caused by one call with my long distance partner who needed to vent.

With Airplane mode enabled the phone managed to crawl over the finish line of the third morning and the experiment was stopped around 0800. Not much else to report here, could have gone longer but did not want to run the battery down and risk it dying if I needed it.

Conclusion

I'm considering this a success as the phone outlasted the manufacturer's specified battery life of (up to) 30h by at least 15h. It's possible that the interventions taken in this experiment could be an accessible way that an existing devices lifespan could be extended without the need for flashing radically new opererating systems*.

An area of possible improvement here is to shift towards more asyncronous, offline first communication and only turn on radio communication at set points in a day. This should both improve battery life and allow for more intentional usage of the device. While most apps and protocols allow for asyncronous usage due to having some level of centralisation, offline first is rare in many implementations and in the case of phone calls, both asyncronous and offline first is imppossible.

Notes

It would be interesting to see if someone running this experiment on OEM versions of android running full GApps sees a similar increase in battery life.