How does Amber for Batteries work?

Amber for Batteries combines Amber’s dynamic wholesale pricing and our SmartShift battery optimisation tech.

SmartShift boosts the value you get from your solar and battery with smart automation that continually hunts for the best prices when charging from or exporting to the grid. This can help you get the same feed-in tariffs that big fossil fuel generators earn (wholesale FiTs that can reach $19/kWh), while still keeping you in control of your battery.

SmartShift works by looking at three key factors:

• Your forecast energy usage in your home
• Your forecast solar generation
• Forecast wholesale prices

These forecasts are combined together to generate a plan for your battery. Read below for more information on how this works.

SmartShift generates a personalised plan for your site

SmartShift models how devices at your home are connected and how power can flow between them. It then uses this model to figure out a plan for your battery (e.g. when to buy from or sell to the grid) that minimises your energy costs over a 24 hour planning horizon.

SmartShift uses the plan to figure out what control actions should be sent to your battery. If the price is low now and forecast to be high later SmartShift might charge your battery so you can either consume that cheap energy yourself, or sell it back to the grid, when prices are higher.

Forecasts affect SmartShift's decisions

SmartShift relies on forecasts when generating a plan for your battery. If high prices are predicted in the future then it might make sense to charge now. However, these future prices aren’t locked in and can change within minutes. This is why SmartShift is constantly recalculating plans as new information about your solar generation, household load, and market prices become available - the algorithm is always trying to make the best decision ‘now’ based on price, load, and solar values predicted in future intervals. The plan screen in the app shows you what SmartShift intends to do based on the latest set of forecasts.

Everything to the left of ‘now’ shows actual household load (blue shaded area), solar generation (yellow shaded area), and prices (pink line) - everything to the right of ‘now’ is a forecast. The dashed green line shows the plan SmartShift has created for battery energy, and the grey shaded area represents times when SmartShift intends to send a command to your battery. In this example SmartShift intends to charge the battery around midday so the household can avoid consuming from the grid in the evening.

Why doesn't my battery always charge overnight?

We might think that prices will be low overnight and high in the morning, but that’s not always the case. If forecasts indicate there won’t be a large morning peak SmartShift might not charge the battery overnight. This is because charging and discharging efficiencies are accounted for when creating a plan. To add 1 kWh of energy to the battery more than 1 kWh needs to consumed from solar or the grid, and to supply 1 kWh to household appliances more than 1 kWh needs to be removed from the battery. These inefficiencies mean charging the battery overnight may not result in the lowest-cost plan if prices are likely to be relatively flat overnight and in the morning.

If there’s a small morning peak SmartShift might want to charge the battery just enough to get through the morning peak without having to consume from the grid. The battery then fills up from solar.

Sometimes high prices don’t end up materialising so the decision to charge can be unnecessary in hindsight. Even so, the decision to charge is sometimes a bit like an insurance policy: a lot of value is obtained from price spikes, so missing out on one, or being stuck with an empty battery during a spike, can be a costly mistake. SmartShift doesn’t have a crystal ball 🔮 so forecasts will never be 100% accurate and not all decisions will be perfect.

The animation below shows what SmartShift was 'seeing' at different points of time for a particular site in April 2025.

SmartShift is capable of delivering significant savings over time

The blue line in the following figure shows the total energy cost over time for a customer in NSW with a 13 kWh battery and 6 kW solar system. The orange line estimates the customer’s total energy cost if SmartShift wasn’t active. Over the course of the year SmartShift unlocked around $1000 of value. This was for the period 1 June 2024 - 1 June 2025.

Three forecasts used by SmartShift

SmartShift uses three forecasts: Price, solar generation and household energy consumption

Price forecasts

Our price forecasting system was developed in collaboration with the CSIRO and uses thousands of data points describing the state of the electricity system to predict prices in future intervals.

Solar forecasts

SmartShift’s solar forecasting system fetches solar irradiance data for a fixed set of points around the country. We then use your historical solar generation data to estimate your solar system size. Combining this information with an irradiance forecast enables us to estimate your solar system’s power output over the next 24 hours.

Household energy consumption

SmartShift uses your historical electricity consumption data to predict household load. Please note it takes time for our load forecasting models to learn your energy consumption habits. When you first sign up SmartShift may use a simplified forecast that assumes household energy consumption today will be similar to yesterday. This works well when consumption is similar from one day to the next but can fall short in certain scenarios. For example, if you charged your EV yesterday the simplified load forecast would assume that you’ll charge your EV again today. After 30 days SmartShift has enough data to start predicting your daily, and even weekly, energy consumption habits.

How SmartShift sends control signals to your battery

At the most fundamental level your battery is controlled by a Battery Energy Management System (BEMS) that exists on your physical device. This system was developed by the battery manufacturer and is generally designed to minimise your household’s interaction with the grid. Because these decisions are made locally adjustments happen very frequently (within seconds).

SmartShift uses a different mechanism to control your battery. Rather than making decisions on the device itself, control signals are sent remotely which override targets set by the BEMS. This mechanism is slower than the BEMS - it can take 10s to several minutes for a remote control signal to reach your battery.

This is why when SmartShift creates a plan it checks if the power target produced by the plan is close to what your battery would otherwise do when operating in it’s default state (where power targets are generated by the BEMS). If it’s close then there’s no need for SmartShift to send a control signal as the BEMS is better at managing sudden changes in load / solar generation. If there’s a big difference then SmartShift sends a control signal to your battery which overrides the default BEMS signal.

Why doesn’t my battery respond immediately?

While we do our best to get commands out to your device as quickly as possible there are often constraints on how fast a device can respond. Some battery manufacturers may restrict how frequently we can send targets to devices, while others may just take a bit longer to act on external control signals. So if you see a delay after issuing a manual command this is normal.

Can I use time-based charging with SmartShift?

Some batteries provide the option of charging during fixed periods throughout the day. These control decisions are also made by the BEMS and are not controlled by SmartShift. It’s generally best to turn this mode off if you’re using SmartShift as the two systems may compete against one other. Time-based control settings don’t account for wholesale prices so there’s a risk your battery charges when prices are high.