Metro de Madrid, in conjunction with Accenture, has developed and implemented a self-learning AI-based ventilation system that minimizes energy costs and emissions and ensures high air quality in metro stations and commuters’ comfort.The artificial intelligence (AI)-based system has enabled Metro de Madrid to reduce its energy costs for ventilation by 25% and cut CO2 emissions by 1,800 tons annually.
On average, 2.3 million commuters use Metro de Madrid’s network of 294 kilometers of track and 301 stations every day. To help passengers stay cool inside stations, particularly during the hot summer months, Metro de Madrid operates 891 ventilation fans, which were consuming as much as 80 gigawatt hours of energy annually.
The Madrid Metro Ventilation experts worked with Accenture Applied Intelligence to develop a system that took inspiration from an unusual source: the coordinated foraging behavior of a bee colony. The system deploys an optimization algorithm that leverages vast amounts of data to explore every possible combination of air temperature, station architecture, train frequency, passenger load and electricity price throughout the day. The algorithm uses both historic and simulated data, factoring in outside and below-ground temperatures over the next 72 hours. Because the algorithm uses machine learning, the system gets better at predicting the optimal balance for each station on the network over time.
The system also includes a simulation engine and maintenance module, which allows for, among other things, tracking for failures in the fans’ operation. This enables Metro de Madrid to easily monitor and manage energy consumption, identify and respond to system deficiencies and proactively conduct equipment maintenance.
The 4th Annual Environmental Leader & Energy Manager Conference takes place May 13 – 15, 2019 in Denver. Learn more here.
