The future of Battery Management system: Adaptive & Personalised management

Lalit Dixit
5 min readApr 2, 2022

--

This is the latest gimmick played by Google when they launched Android 12. Thus, if you have bought any android phone in recent times, you might have seen an icon for enabling adaptive battery in your phone settings. If you are an iPhone user, the chances are you are already using it.

It is the short term for adaptive battery management. The idea behind the technology is to learn from a user’s behaviour and control the battery discharging based on these learnings. It relies a lot on AI & ML.

Though not pronounced, it was already researched in almost all sorts of devices. But Google decided to coin the term “adaptive battery” as they doubled their efforts to personalize the phone experience for each user. It fits into their long-term vision of providing a personalized experience to each user. After all, a phone is an extension of a user’s personality and should provide a unique user experience to its users.

What is Adaptive Battery?

It is the short term for adaptive battery management. The idea behind the technology is to learn from a user’s behaviour and control the battery discharging based on these learnings. It relies a lot on AI & ML.

While the earlier battery management systems were designed to optimize the battery discharging to prolong daily battery life. They did not focus on user behaviour so much.

To understand it better, think about a user who is accessing Facebook while travelling. The phone is in motion with cellular connectivity changing very fast. The constant connection attempts consume a lot of energy. As a result, after 1–2 hours of usage, the phone battery would become low. The software doesn’t have many options in this case. It can either switch to a connection that is stable and doesn’t offer much internet speed and thus generate a subpar user experience, or it can keep switching networks to connect with a high-speed network and end up with a low battery after 2 hours.

Battery in absence of adaptive BMS

Introduction of personalized battery management

As smartphones are growing with new chips that are capable of handling multiple activities and neural networking capabilities. The phones now have the capabilities to learn from the user requirements.

The new devices log user actions & responses and learn from them. Thus, when the user opts for adaptive battery management, the phone starts learning from the user’s behaviour. It notes when the user is engaging with the phone, what kind of applications he usually loads, how much time does he spend on each application, what are the popular times when he charges his phone, how much does he travel, which networks fall into the travelling region, how much internet speed is needed for the actions when the user travels and a lot more other things that control the user experience.

In a sense, the phone is noting each activity done by the user and generates a user profile that includes the user preferences. Based on the learnings, the phone starts controlling the background power supply.

An adaptive BMS balances the charge distribution to all functions based on its learnings

For the same example that we discussed earlier, now the phone relies on a lot of data. The phone already knows that the user has a habit of accessing Facebook while travelling. It also knows if messaging is the primary need or if the user is only interested in the newsfeed during the travel. Thus, when the user starts travelling, the phone doesn’t sync emails and contacts but only uses energy for syncing Facebook app. Also, it loads most of the data when the networks are stable which results in less energy consumption. It loads a lot of high-quality newsfeed content when the user is not in motion and keeps it stored to showcase when the user starts traveling. For the remaining time, it just keeps the user connected.

The higher the usage, the more the phone learns from usage patterns and optimizes the battery

How does adaptive battery management fare with respect to normal BMS?

Adaptive and personalized battery management is better than the normal BMS. The system fails only when the device is being used by multiple users or the behavior of the user keeps on changing.

Also, since no human behaves in a fixed way. We keep on changing our behaviours and thus, the phones keep on failing. Only when one uses the system for some time, the phone becomes good enough in managing the battery based on its learnings.

What are the issues with personalized approaches?

The biggest flaw is the risk of data theft and there are other privacy concerns. A personalized BMS requires the system to observe every activity of the user. A highly adaptive version would store each action, location history, user interaction, and user preference to control the battery usage. But, when a device collects so much data, it also creates a risk of unfair usage of that data.

These learnings can be used to target users, manipulate their thoughts, change their opinions, and can also be used by malicious organizations & individuals to hurt the owners.

What’s the future of personalized BMS?

It is going to stay and thrive. AI & ML is revolutionizing the domain and the benefits are very high. You can expect more user controls in the coming days.

Asking for users’ permission to collect their behaviour stats is a good step in this direction.

Further advancements can include storing user data only on the phone, encrypting the data, not associating user’s identifiable information to any file, using virtual IDs to identify users, and providing an option to delete the data whenever any user wishes.

--

--

Lalit Dixit
Lalit Dixit

Written by Lalit Dixit

In a complicated world full of random data, I exist to uncomplicate

No responses yet