Andy Jassy | President and CEO of Amazon.com | Amazon website
Amazon is leveraging artificial intelligence to enhance the sustainability of its buildings worldwide. In Glasgow, Scotland, a faulty valve in an underground pipeline was detected by Amazon's AI-powered utility efficiency tool, FlowMS. The system, built using Amazon Web Services (AWS), identified an anomaly in water usage and alerted employees. As a result, engineers discovered and repaired the leak, preventing the loss of 9 million gallons of water annually.
Kara Hurst, Amazon’s chief sustainability officer, stated: “At Amazon, we’re innovating with AI to help us find new ways to decarbonize even faster, including inventing new solutions that continue to make our buildings more energy- and water-efficient.” She emphasized how these innovations are part of Amazon's broader efforts to improve efficiency across its operations.
Amazon's Decision Science & Technology team has also developed the Base Building Advanced Monitoring (BBAM) tool. This system monitors HVAC systems using AWS machine learning tools such as Amazon SageMaker and Lambda. It analyzes operational data to identify issues like clogged filters or unexpected energy consumption due to weather changes. Sites utilizing BBAM have reported significant energy savings by addressing problems proactively.
In New York, FlowMS helped discover a miscalibrated utility meter that inaccurately reported high energy use at a fulfillment center. Similarly, in Spain, BBAM identified a malfunctioning air conditioning unit by analyzing cooling output against expected demand based on weather conditions.
The company plans to expand FlowMS and BBAM tools from their current deployment at 120 sites to over 300 buildings globally by the end of 2025.
Additionally, Amazon has developed Advanced Refrigeration Monitoring (ARM) for real-time analysis of refrigeration units at fulfillment centers. ARM helps maintain optimal temperatures for perishable goods by predicting potential operational issues through data analysis. At one site in Spain, ARM detected equipment faults early enough to prevent food loss and reduce downtime significantly.