“AI-supported tools Amazon uses to monitor utility and building system performance in 120 of its sites globally will be expanded to more than 300 facilities by the end of the year,” reports Utility Dive. “The tools have helped identify water leaks, faulty energy meters and malfunctioning refrigeration equipment at its facilities….Amazon is using the technology to analyze HVAC operational data and energy consumption, monitor refrigeration units, identify issues like leaks or clogged filters,” hoping to save money and reduce carbon emissions.
Artificial intelligence tools “have already helped Amazon reduce water and energy consumption at sites in Scotland, Spain and New York.” For example, in Amazon’s Scotland facility, “the FlowMS tool helped associates uncover a leak in an underground water line that could have caused up to 9 million gallons in annual water losses…The tool analyzed the building’s water meter data, saw that it was using more water than expected and alerted engineers, who traced the leak and repaired a faulty valve.”
The Base Building AI tool makes use of Amazon’s SageMaker and Lambda machine learning capabilities to detect system anomalies by analyzing HVAC operational data, energy consumption and weather data. “The tool helped Amazon pinpoint a miscalibrated utility meter at a New York facility that appeared to be using five times more energy than nearby facilities. Separately, Amazon identified a malfunctioning air conditioning unit at a Spanish facility by comparing its actual cooling output in real time to expected performance based on outside weather conditions.”
A refrigeration-monitoring tool “analyzes fulfillment-center refrigeration units in real time to help maintain ideal temperatures for perishable goods. The tool predicts which system components are involved in a possible anomaly and alerts company associates so they can take action. At a site in Spain, the refrigeration tool identified a change in the defrost cycle caused by a piece of malfunctioning equipment, preventing ‘significant food loss and an estimated 1,000 hours of equipment downtime.’”
Artificial intelligence spotted a wildfire in California’s Orange County and alerted firefighters, who put it out.
Some robots using artificial intelligence can identify and kill 100,000 weeds per hour.
Scientists are using artificial intelligence to identity some of the trillions of viruses that live inside human beings.
Viruses can be useful. Scientists have engineered a virus to steal proteins from the HIV virus, potentially eliminating AIDS.
A virus is being used to cure deafness in new gene therapy. Researchers also discovered that a plant virus could be used to save crops from root-eating pests.
Genetically-engineered skin bacteria keep some mosquitoes away. Scientists have genetically engineered a fly species to eat more waste.
Robots with artificial intelligence will reduce the need for weed-killer and pesticides by more precisely targeting weeds and pests. That will cut farmers’ costs, and radically reduce the size of the crop chemical industry, because robots will use up to 90% less spray to kill the same number of weeds and pests.
On some American farms, there are already drones with artificial intelligence that spray fungicides to kill pests. As Bloomberg News notes, “These aerial acrobats use less than a tenth of the energy of ground tractors — and they don’t squash the crops, rut the earth or even touch the soil.”
Robots with artificial intelligence are spreading on Japanese farms.
