Buildings, especially large ones such as factories, warehouses, data centers, and office buildings consume a large amount of electricity. That electricity costs a lot. Whole building energy systems, including HVAC systems, are often complex. That complexity compounds as distributed energy resources (DERs, such as connected rooftop solar panels, batteries, and electric vehicles) are connected. That compounding is magnified when buildings, like the ones mentioned above, are connected together into systems of buildings (including microgrids). Making everything more complex, and often more costly, is that the rate for electricity can vary significantly if the building is subject to time dynamic rates. High costs and complexity can make for a big operational headache that distract from business objectives.
At yize nrg (pronounced “wise energy”), we make it easier and more affordable to operate these large buildings by making their energy systems intelligent. To make them intelligent, we use intelligent agents, known as “Brainy”, that optimize operations and lower electricity use, especially when electricity costs are high.
We can operate in a way where your building occupants barely notice. We thrive on complexity. More complexity means more opportunity for cost savings. Are your rates connected to wholesale market rates? Do you have rooftop solar assets? Do you have a complex of buildings? Do you have a microgrid? Great; that means that we can save you even more money. This concept, using AI agents to significantly reduce energy costs, has been proven to work (Google saved a whopping 40% on data center cooling costs).
Our intelligent control system, Brainy, can reach electricity cost savings of 50% in high complexity, high flexibility environments.
Our system of intelligent agents (AIs) are collectively and individually known as “Brainy”. These Brainy agents are stable, reliable, and based on cutting-edge deep reinforcement learning techniques. These agents deliver increasing savings as the complexity increases.
Brainy trains, constantly improving itself, in a state-of-the-art virtual environment that can model everything from simple single-family homes to highly-complex systems of factory buildings equipped with high penetrations of DERs (ranging from flexible EV units to rooftop solar). We are able to model microgrids as well as traditional demand response (DR) methods. This environment is based on years of US Department of Energy research that is freely available.
We use a variety of techniques to constantly improve Brainy agents. We have Brainy agents that actively design new and better architectures of other Brainy agents in a process called meta learning.
On the low end, Brainy consistently delivers 10% in electricity cost savings in environments with very low flexibility, and can deliver 50% in electricity cost savings in environments of high complexity and high flexibility (including microgrids).
We are tackling the large challenges posed by the electricity grid becoming more distributed as increasing amounts of Distributed Energy Resources (DERs, such as connected HVAC systems, rooftop solar panels, batteries, and electric vehicles) and also posed by its transition to non carbon-based energy sources.
The nrg Platform is our vision. It is in development and will be available well after our intelligent building energy systems. We do not yet have a timeline for its public release.
To solve the challenges facing the energy grid, we are building a platform to facilitate automated communication between the buildings, the DERs, the distribution utility, and other stakeholders connected to an electricity distribution grid. We are developing intelligent agents, various distributions of Brainy (our robust “AI” system), to integrate with the platform to coordinate communications as well as optimizing the operations of the connected buildings and devices.
The nrg platform is still in its early stages but is developing rapidly. The platform, with its intriguing architecture, is the product of years of iterations and working with stakeholders across the energy industry. One example of its novel architecture is that it radically evolves the concept of virtual power plants (VPPs). Please contact us if you would like to learn more about it.
We are an experienced team with backgrounds in electric power engineering, software engineering, and machine learning. We build safe, reliable, high-performance systems that are on the cutting edge. We are based in New York.