Smart2: Smart Charging for Smart Phones


In this paper, we present Smart², an advanced smartphone charger that mitigates battery’s capacity fading, which until now has usually been ignored. Smart² exploits the fact that many users charge their phones over night. Since the overnight charging duration is unnecessarily long, the battery is subjected to a high average state of charge (SOC), which accelerates battery aging. Therefore, we delay the charging adaptively to be done shortly before the phone is unplugged. With this scheme, clearly when averaged over the duration of the night, the average SOC is lower and hence aging is reduced. Indicators are a set alarm clock and/or statistics of previous usage. Similarly, we lower the maximum target SOC. To enable this, the main challenges are firstly to find a solution that does not negatively influence the usability and secondly to quantify the achieved savings in terms of aging mitigation. Towards this, we propose a novel charging scheme which can be implemented in the smartphone’s firmware. Furthermore, we propose a modified battery charging device that can be used with almost all existing smart phone models. Using our proposed techniques, the average battery cycle life can be nearly doubled from 3.7 to 6.6 years.

IEEE 21st International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA)


Why smart charging?

The battery as an energy storage is one of the most important parts of mobile devices because the capacity of a battery and the power consumption of a device determines its runtime. However, the capacity of a battery fades over time due to aging and as a result the runtime decreases too. Some devices are equipped with non-removable batteries which means that the lifetime of a device will end with the death of its battery.

Many factors are known that can prolong the battery lifetime and there are guides available how to treat a battery correctly. However, most users do not want to care about their battery lifetime or even adapt their behavior to extend it. As a solution, technical methods for intelligent charging have to act in the background without disturbing the user.

Therefore, I have programmed an Android application for smartphones that is capable of controlling and monitoring the complete charging process on condition that the hardware of the smartphone provides the necessary interfaces. It starts automatically if a power source is detected and terminates after charging is finished. The user can adjust timing and threshold values as well as control the charging process manually.

Concepts to reduce Battery Aging

Lowering Storage SOC

Most people charge their smartphone over night. In the evening, when it is plugged into the charger it immediately begins to charge. When the SOC reaches $100%$ the cell remains at this level for the rest of the night. This scenario is not ideal for the battery since life-time is reduced if the SOC stays at $100%$ for a long period of time. One idea to slow down battery aging is to keep the SOC at a lower level if it is possible. For example the battery could only be charged to $50%$, and remain at this level. Only shortly before the smartphone is unplugged, the battery is fully charged to $100%$.

Lowering storage SOC.

Lowering Maximum Battery Voltage

Another idea to prevent battery aging is lowering the maximum battery voltage. For lithium-ion batteries the default maximum voltage in smartphones is $\SI{4.2}{\volt}$. As a rule of thumb, every reduction of $\SI{0.1}{\volt}$ of the constant charge voltage doubles the cycle life. The charging voltage is set by the power management chip inside the smartphone. Although many power management chips can adjust the charging voltage, the Android API does not provide a function to alter this value. However, a lower maximum voltage corresponds to a lower maximum SOC. Therefore the only way to influence the maximum voltage is to monitor the SOC while charging and stop charging at a lower value than 100%. This will also cause the open circuit voltage of the battery to stop at a lower level. Of course, the charging voltage will still rise to $\SI{4.2}{\volt}$ in the charging phase of constant voltage but if charging is stopped before the battery is full, the voltage of the battery will drop back below $\SI{4.2}{\volt}$.

Lowering the maximum SOC.

Lower $\Delta \mathrm{SOC}$ with Partial Charging

The $\Delta \mathrm{SOC}$ has a great impact on cycle life. The cycle life can be multiplied if the $\Delta \mathrm{SOC}$ is kept small. Since the energy required during the day is determined by phone usage, the $\Delta \mathrm{SOC}$ can only be lowered by either operating the phone more power efficient—which is task of the power management—or by charging the battery more often which will of course increase the cycle rate. The question is, if the cycle life is increased more than the cycling rate increases. Fact is, that there are many cases where the battery is charged more often by the user than once per day. The battery will also be charged during the day if it would not last for the whole day or if the user connects the phone to a PC/notebook for data transfer. However if the battery is fully charged during the day, the battery will be probably charged more than required. Additionally, the SOC at the end of the day could also be higher than the optimum. To improve this behavior, the battery should only be charged partially to a value that guarantees enough energy for operating the phone for the rest of the day but keeps the SOC nearer to the optimal value.

Partial Charging.