Hi. My name is Yael Grushka-Cockayne. I'm an Associate Professor at the Darden School of Business, one of Mike's colleagues. I'm here to tell you about a Big Data transformation study that I conducted with a team of coauthors with Heathrow Airport and Eurocontrol. Eurocontrol is the body in charge of the European airspace, similar to the FAA in the US. Over the years, Eurocontrol have had quite a big challenge on their hands. All the different systems in Europe related to the airspace are quite fragmented. Different countries have different airports and airlines that operate in different languages and use different systems. Eurocontrols has a vision. Their vision is for a single European sky for 2025. They are trying to improve the efficiency, the safety, and the flexibility of the system across the entire airspace. In order to do that, one of their initiatives is to work with the airports to form what's called an operations center, or the acronym as an airport operations center or APOC. The vision is that if each airport positioned all of the different stakeholders in one big room, having all of their screens and their data in one central location and the key decision makers all in that room, there could be huge efficiency gains to the airport and to all the different stakeholders and players in the airspace. One of the first APOCs, one of the first operation centers that were set up in Europe was in Heathrow Airport. Heathrow Airport, just southwest of London, is one of the busiest airports in Europe and in the world in general. It has over 75 million passengers a year, five terminals, two runways, and flying to 250 destinations. Twenty four million of those passengers, transferring passengers. Passengers that land at Heathrow, but then fly out of Heathrow to a different destination. The vision that Eurocontrol had, which they approached myself and my team, was a vision around how Big Data got used in the APOC, in this center with all the different stakeholders. Their vision for the Big Data was that we could use the data that got stored in all the different systems, the systems that the security managers had, the systems that the airlines, that the baggage control systems had. If I could coordinate and look at all the different data, and we could look at it in real-time, could there be efficiency gains for how the stakeholders interacted and how decision making was made? Ultimately, the goal of Eurocontrol and of Heathrow specifically, was happier passengers that all find their bags, that arrive and depart on time. And the big question in the vision was, could we use the data and Big Data to achieve that? Once we got onboard and we started to have these conversations, we knew that in order to nail it down and to actually get to something that we could implement, we needed to identify how data got used. What was the usage of the data across the organization, across the entire airport? Well, data gets used in many different ways at Heathrow. Data gets used to help with baggage flow. Data gets used to help allocate airplanes to different stands depending on when they arrive, where they're coming from, and where they're headed towards. Data gets used to help with passengers flow across the organization in the airport. Specifically, we were curious to see how we could use data to better anticipate a better forecast, the connection journey or the connecting journey of transferring passengers. Could we use data from the various systems to understand how many passengers are going to be connecting at a given moment, how much time it's going to take them to flow through the airport, will they miss their connecting flights, and are we going to have problems at the security lines with big number of passengers arriving at the same time? And so, we've focused our usage and our attention for this project on how data gets used and how data can transform the actual connecting journey of passengers at Heathrow Airport. Once we identified that that was the usage we were focusing on, we then started to break down the different components of the engines. What are the capabilities that we need to identify? Where does the technologies lie? What technology gets used? What data gets stored? And is this data that can be useful in thinking about these transferring passengers? Who are the people and the decision makers and the organizations involved along the process? For instance, some of the people could be the people in charge of the security lines related to immigration. Some of the people involved in the organizations involved with all those related to the transportation within the airport, such as buses and the different trolleys that sometimes help you nip around the airport if you're really in a hurry. We had to identify the processes. And in this case, it was the passenger journey from A to B. What does a passenger do when they land, all the way to the next gate as they're transferring through Heathrow Airport? What are the different checkpoints that they hit? And could we identify opportunities for improvement and opportunities for improved flow at each one of those points throughout the process? That was the engine. Well, we did this, and this was the bulk of our time, and the bulk of this project was really breaking it down and identifying that there are actually some struggles and some non-trivial components to this engine. Data gets stored in many different systems. Over six or seven different databases house pieces of the information that we needed to track a passenger from when it's still in the air, on an airplane approaching the airport, to when it gets to its final gate and are about to board for its destination. Tracking a passenger throughout the process and making sure that they actually arrive where they need to be on time was a hugely hard task. And we realize that the systems had different information that didn't always either agree, meaning inconsistencies, or we didn't always have the information in real-time the way that we needed it on the spot to make decisions. It turned out that the key or the secret to our success and the key to helping us crack this project was actually in the baggage information. The baggage detail and the data that got stored in the baggage handling system related individual passenger to a seat on an inbound flight and told us something about their destination and helped us track the progress and translate the different information from one system to the next. We used the information. We used over 3.7 million records of passengers coming into Heathrow, and we came up with a model that told us something new about what predicts and what influences a transferring passenger's connection time. For instance, and some of you may know this actually intuitively, it turns out that where you sit on an airplane, what class your ticket is, and how close you are to the front affects how long it will take you to get through the process. It also makes a difference what kind of airplane you're on, if it's a small or a wide airplane. Where you get to stand in the airport? How often have you actually been on an airplane? You see that you're running low on time, and you tell yourself, with my luck, my airplane is going to get to the furthest gate, and I'm going to have to run all the way to the other end of the airport. Well, we can prove, and we proved with our model that this is true, and that where the airplane get on, what stand it gets, what gate it gets allocated to, where it's coming from, where the passenger is seated, if a passenger is traveling on their own or with a group, all of these are factors that affect the connection time. Using all of this information, we can build models, we can build different profiles for passengers and put them into a big engine that allowed the decision makers to come up with predictions as to how many passengers are going to hit different key points within the airport. As we step back, we did some testing, we did some live trials, and we recognized, and we proved to ourself that our system is fairly good, is coming up with predictive distributions and telling us something about those connection times. We stepped back, and we reflected. And we told ourselves, what is required from the entire APOC, and more than that, the entire ecosystem around the airport activity? What is required for this data to be utilized on a regular basis? What kind of data ecosystem are we looking for, are we in need for, in order for this system in our model to be used on a regular basis and for the decisions to be improved and for the decision to rely on this data on a regular basis? The ecosystem had to do with the different stakeholders. We had to go back to the airlines. We had to talk with them about their processes around real-time data sharing. We had to encourage them to be more transparent with each other, to go back even to the airplane manufacturers to come up with software systems that would support this type of real-time communication, to store the data better, and to be open and transparent with all the different stakeholders in a park, and to coordinate upfront as to how the data is going to get recorded in order for us to feed it into our model and then use it on a regular basis. What I saw that day at Heathrow and what I've been seeing since and throughout my conversations with them is there is a lot of hope and a lot of belief that this use of data will transform the way that they operate and communicate with each other. No longer do they see themselves as a fragmented set of stakeholders that are being forced to share a single room. They are actually under the impression, and they believe that they are now operating as one with the idea that data gets tracked throughout the process, and their stakeholders can follow it and track it throughout. And who will benefit from this? Everybody, from the passenger, to the airlines, the airports and globally, the entire airspace.