When people think of AI, the first thing that comes to mind is disastrous images of self-aware robots trying to destroy humanity and take over the world, just like in many hit movies. They couldn’t be more wrong!
Artificial Intelligence is designed to mimic the human way of thinking. AI software therefore has the ability to think and act independently. In building automation, this means that AI can make a significant contribution to improving maintenance, comfort and energy savings.
A year ago, AKKA Netherlands started working on an internal project focusing on how AI can help our clients reduce their buildings’ energy consumption.
At this point in time, green energy sources are not able to power our entire planet. That is why it is essential that we use and store the energy we do have in a smart way. Artificial Intelligence is perfect to take care of this task. The technology is self-learning and works independently, without any human intervention. By integrating AI into building management, you can « teach » your building to find out for itself what are the best renewable energy sources to use, store and generate.
In current standard building management systems a “control system” is used. Broadly speaking, the goal of a control system is to determine the correct inputs (actions) into a system that will generate the desired system behavior.
With feedback control systems, the controller uses state observations to improve performance and correct for random disturbances and errors. Building management systems use that feedback, along with a model of the plant and environment, to design the controller to meet the system requirements. This concept is simple to put into words, but it can quickly become difficult to achieve when the system is hard to model, is highly nonlinear, or has large state and action spaces.
The goal of AI is similar to that of the control system – it just uses a different approach and different terms to represent the same concepts. Both methods want to determine the correct inputs for a system that will generate a desired system behavior.
Over the last decade, we have learned that the use of AI to solve the « control system problem » is more effective and powerful than traditional control systems solutions.
Using AI in buildings is like driving a self-driving car. You simply choose your destination and he car will take you there as quickly and safely as possible. That is also the case for AI driven buildings. You enter your preferences into the system and the building finds out the best renewable energy sources to use.
The main goal of our algorithm is to use AI to reduce energy consumption. In our investigation we are also looking into how we can optimize the comfort level in the building and at the same time reduces energy consumption.
The idea is that the algorithm looks ahead 24 hours, calculating scenarios based on combinations of most likely events. This way, the algorithm learns about the behavior of the building and about the influence of factors such as the changing supply of sustainable energy sources or weather conditions, such as outside air temperature, sun and wind. With this information, the algorithm then makes smarter use of climate installations and available energy sources in the building: renewable and sustainable energy sources can be used before other sources are considered. Moreover, the system will take into account the energy prices. Ultimately, the software controls the installations in such a way that the optimal climate is achieved with the lowest possible use of energy. All of this surely results in a reduced use of energy and consequently, savings on energy bills.
So no, AI is not going to take over the world and there is absolutely no question of malicious robots. On the contrary, with self-learning software you can save energy without being actively involved yourself. We can therefore say that you can contribute to saving the planet from the comfort of your home. The truth is that technology is neither good nor bad. It all depends on how you use it and what you use it for.
Author of the article
CTO AKKA Netherlands