Votes: 0
Views: 1001

rAider (Refrigeration Artificial Intelligence DEmand Response) is a palm-sized IoT/controls package that uses big data and artificial intelligence with sensors to optimally control commercial refrigeration for food stores and the power grid. First, our IoT device collects refrigeration and environmental data from stores using sensors. The data wirelessly transmits to our cloud data storage. Our servers pull this data and use it to train our machine learning (ML) algorithm, finding each refrigeration unit’s optimal settings. The IoT device then wirelessly receives these optimal commands from the algorithm. The device’s tiny central processing unit (CPU) serves as the brain for this non-invasive automation solution and provides energy efficiency and demand response (DR) benefits for the commercial refrigeration and freezer systems across each store. DR is the service of shifting loads away from high-demand times of day, e.g., 4-9 PM when homes typically see peak demand. rAider shifts refrigeration peak power outside this time to reduce aggregate demand across the grid and in turn reduce pollution from power generation.

The U.S. alone has over 200,000 fresh food stores. These stores are relying more and more on refrigeration to compete against ecommerce rivals like Amazon and Walmart. This pressure holds food stores to low profit margins of 1%-2% in the U.S. At the same time, the stores face growing costs with ever-increasing pressure from customer expectations for food quality and low prices. As investment in commercial refrigeration increases by 6.2% annually across the U.S., these refrigeration systems constitute the largest operational cost and biggest energy consumer in food retail stores (44%-62% of their electrical energy). Commercial refrigeration’s power also doubles 3-4 times each day with power spikes to defrost the system. With operational and energy data from commercial appliances, rAider balances these power spikes and reduces store energy bills by 17%-47%. It continues to improve and provide more savings over time, since it constantly improves our neural network to find hidden correlations. By implementing the control settings from our algorithm for defrost and compressor cycles across each store, rAider simultaneously provides savings for retailers (through DR rebates from the grid and reduced energy consumption).

rAider will reduce energy consumed in each food store by an average of 174 MWh annually. At the same time, the technology will lower peak demand and greenhouse gas (GHG) emissions across the power grid by reducing fuel-burning generation. According to the U.S. EPA's GHG Equivalency Calculator, the carbon offset will be equivalent to reducing carbon dioxide emissions by 24.6 billion metric tons across the U.S. every year. Altogether the total addressable U.S. market for rAider is over $24 billion/year. In summary, our IoT device optimizes how commercial refrigeration operates to reduce energy waste and costs for food stores and their electrical utilities.



Voting is closed!


  • Name:
    Jesse Thornburg
  • Type of entry:
    Team members:
    Dr. Jesse Thornburg
    Dr. Javad Mohammadi
  • Profession:
  • Number of times previously entering contest:
  • Jesse's favorite design and analysis tools:
    MATLAB, Simulink, EnergyPlus
  • Jesse is inspired by:
    Nature's patterns and textures inspire my designs. I have always been a tinkerer and a builder, so I love a systems approach to engineering: designing, building, testing, and debugging. I love to assemble and work with multi-disciplinary teams, and their diverse skills inspire me. Through growing up on a family farm, and there becoming used to constant problem solving, I appreciate seeing elegant solutions that are aesthetic, functional, and simply carried out.
  • Software used for this entry:
    EnergyPlus and Modelica
  • Patent status: