Business Case: London Heathrow Airport Launches BI and Machine Learning to Improve Airfield Management, Predict Passenger Flow, and Transform Airport Security Heathrow airport in London is the second busiest international airport in the world, second only to Dubai international airport in number of airplanes landing and taking off each day and the seventh largest in terms of total passenger traffic. Managing over 215,000 passengers every day is a challenging task and requires a high degree of coordination to manage passenger traffic and give passengers a smooth airport experience. Any unexpected disruptions in the smooth workflow in operations at Heathrow such as damaged runways, storms, delayed or canceled flights, shifts in jet streams, etc. would disturb the entire functioning of the airport, passengers and airport employees. Data analysts at London Heathrow were using Excel spreadsheets to analyze its airfield, passenger and flight data and sorely needed a centralized management system that would extract large volumes of data produced by airport operations and transform them into useful visual insights. Stuart Birrell, CIO at Heathrow was concerned that We have tens of thousands of people who work around the airfield. Safety is critical. Adopting tools like Power BI makes life easier. It is the simple things. There is GPS in the airfield vehicles. If a driver finds a problem with the concrete, this can be recorded accurately. Heathrow chose Microsoft Power BI as their BI solution. The reporting produced by its BI tool ensured airfield safety, allowed airport staff to function better and improved passenger management. The key was moving from a paper-based, reactive operations model to a more predictive, proactive planning model in which staff were dealt fewer surprises on a day-to-day basis that enabled them to change their plans on- the-fly. The answer was BI reports and dashboards that were made available to airfield managers, security officers, transfers and customer service staff and a machine learning model that accurately predicts passenger flow in 15-minute increments into each terminal. Birrell says it's possible to mash up historical scheduling data and a feedback loop to provide more accurate forecasts. With insights from these data analytics tools managers could plan staff breaks, open and close security lanes as needed and schedule staff shifts to balance passenger flow across the airport in peak times. As Birrell said, For passengers, it is all about getting them to aircraft on time. The new system also helps manage arrivals. Under the old model, if several flights came into the airport an hour early because of tailwinds immigration and baggage staff would have to scramble to react to the sudden spike of arriving passengers. After the predictive model was deployed, the airport manager could share the insights with air traffic control and security staff to better schedule immigration and security lanes and teams by knowing wher.