At Electrifai, we’re leveraging the power of AI to create a safer and smarter home environment. ElectrifAi believe that once everyone has access to affordable, sustainable energy products, society can be more creative, productive and joyful.
ElectrifAi are a company born out of the Stanford Clean Energy Prize competition with both hardware and software expertise in electrical engineering, microelectronics design and embedded systems development. We are now ready to start manufacturing the first verified, commercially available solar home energy systems utilizing AI and machine learning technologies.
In this post, we will introduce our vision of a more sustainable society and follow it with a brief technical description of our architecture.
What we are trying to do is create an environment where everyone in the world has access to affordable, ubiquitous solar power through home energy products. This will end the cycle of energy poverty, where 2.7 billion people today have no access to electricity and depend on dangerous and polluting alternatives such as diesel generators.
To achieve this vision, we are building AI into our home energy systems to reduce their cost and make them more convenient for consumers. For example, we are using cameras to recognize when a panel is dirty and send an alert to the owner so that they can easily clean it; or using microphones for predictive maintenance so that customers can schedule replacement parts before something breaks down.
This level of automation is not possible today, but we are confident that AI will play a central role in decreasing the costs and increasing the convenience of such products.
Our AI system consists of a hardware component, which measures solar energy and transmits data to the cloud, and a software component, which uses machine learning to learn patterns in real-time. We have designed this architecture in such a way that we can scale up as our home energy market develops, going from thousands to millions of solar panels around the world.
The hardware component is a physical device that measures solar energy. We are developing state-of-the-art hardware for this purpose and have already begun shipping prototypes for field testing. We are building our prototype on top of a very well known architecture, the Raspberry Pi with its ARM processor, I2C bus and the GPU. This allows us to keep our development cost low and experiment quickly without incurring a large amount of R&D cost or custom PCB design costs.