When should a battery charge? When is it better to discharge? And how can you already respond today to tomorrow’s energy prices, solar generation and consumption peaks? These are all questions AI should be able to answer. But can it really, or is it still mostly hype without tangible value? In this blog, we explain how we use AI in our systems at Smart Ass Energy. Spoiler: it genuinely makes a big impact.
Ask someone what AI is and there is a good chance they will answer “ChatGPT”. Understandably so, as generative AI has become highly visible in recent years. But in energy management, the real value lies elsewhere.
Machine learning, neural networks and autonomous decision-making models make it possible to continuously analyse complex energy systems and independently make the most profitable choices. Not based on pre-programmed rules, but on data, forecasts and experience.
Each connected asset runs a machine learning model that learns the behaviour of that specific battery, EV charger or installation. In the cloud, an AI agent then brings all the information together and determines how the entire energy system can be controlled optimally.
We also use a large language model (LLM) to make it easier for users to interact with the system. This allows text input to be automatically translated into usable system input.
But make no mistake: that chat interface is not what sets our technology apart. The real intelligence lies in the self-learning models that continuously make decisions and keep improving themselves.
Many traditional energy management systems still operate according to fixed logic.
That works, but only as long as reality follows those rules. An AI-driven EMS works fundamentally differently.
Instead of programming hundreds of rules in advance, we only define the boundaries and limits of the different assets. Within those boundaries, the AI model determines which decision makes the most sense at that moment and is financially most attractive. This makes the system much more flexible than traditional rule-based software.
A good decision never depends on a single parameter. The system continuously combines different data sources, such as:
This allows the system to think ahead. Are high electricity prices expected tomorrow? Then it can decide today to store energy.
Are a few hours of very cheap electricity coming up? Then it may make sense to deliberately discharge the battery first, so it can be recharged later at a lower price.
Does the system anticipate a consumption peak or limited grid capacity? Then it can choose to charge the battery as much as possible before that situation occurs.
The result is a much more intelligent form of energy management, where decisions do not just respond to what is happening, but also anticipate what is coming next.
An important part of our technology is machine learning at asset level. Our Load Commander automatically learns how a battery or other asset behaves in practice.
No one needs to configure everything manually. Among other things, the system learns:
This has several advantages. New assets can be integrated faster, without lengthy configuration. In addition, control becomes increasingly accurate as more data becomes available.
And it creates another opportunity: predictive maintenance. If a battery suddenly starts behaving differently than expected, the system can detect and flag the deviation before an actual problem occurs.
All information ultimately comes together in one central AI agent. It combines data from the different assets with external information such as weather forecasts, energy prices, grid constraints and user preferences.
Based on this, it determines how all parts of the energy system should work together. Not just the battery, but also EV chargers, solar panels and other assets are controlled as one integrated system.
It is precisely this central intelligence that makes it possible to continuously find the optimal balance between returns, comfort and energy availability.
It is often assumed that better energy performance automatically requires larger batteries. That is far from always true.
A smart Energy Management System gets more out of the same installation by making better decisions at every moment. As a result, a smaller battery can achieve surprisingly much in many situations.
Curious what our EMS can do for your energy management? Get in touch with our experts. We would be happy to explore how your business can gain more control, more flexibility and a higher return from its energy.