Over the past couple of years, embedded problems have been showing up resulting in acute complexities and hence bringing companies to search for a solution. Google has been on try-out assessment with Algorithm to regulate the cooling systems that include fans, ventilation, and other paraphernalia in order to lower power consumption.
Now Google and DeepMind affirm that it has effectively handed control to the algorithm, which will help in managing cooling at several of the Google data centers. Previously this system was used for placing recommendations to data center managers who would then decide whether to implement Algorithms or not so to save energy around 40 percent.
Now we’re taking this system to the next level, our AI system is directly controlling data center to cool. This first-of-its-kind cloud-based control system is now safely delivering energy savings in multiple Google data centers. – DeepMind, the London-based artificial-intelligence company Google acquired in 2014
This new project will try to exhibit the potentiality for artificial intelligence to manage infrastructure and help to make it appear how advanced AI systems can work in collaboration with humans.
For every five minutes, the Google-based AI will take a snapshot of the Google data center cooling system, helping to predict different combinations of potential actions which can affect the future energy consumption. The AI system then comes in rescue to identify which actions will minimize the energy consumption which is then verified by the local control system so that it can be implemented.
We’ve designed our AI agents and the underlying control infrastructure from the ground up with safety and reliability in mind, and use eight different mechanisms to ensure the system will behave as intended at all times. – DeepMind
The new system has safety controls to prevent it from doing anything that has an adverse effect on cooling.
The issue of energy consumption has gained acute attention in the tech industry. As according to the 2016 report submitted by US Department of Energy’s Lawrence Berkeley National Laboratory, the US data centers has consumed about 70 billion kilowatt-hours in 2014 which is about 1.8 percent of total national electricity use.
It is expected that the new AI control system would be dependable as well as be safe for the deliverance of energy saving. DeepMinds hope that this improved machine learning shall be applied in industrial settings in order to optimize their loads and tackle climate change of even bigger unfolds.