Decision Lab Case Study with DSTL & Department for Transport
Can we understand the impact of different types of security measures in airports?
Complex environment with significant interactions is extremely hard to model.
Our agent-based modelling, trialled on Heathrow Terminal 5’s departure concourse and its daily flow of 30,000 passengers, found an answer.
The impact of security checks on the airport operations
Airport security is a high priority and attack targets include concourses and other public spaces.
The UK government wanted to understand the impact of different types of security checks as people enter the airport terminal building to go check in. This included determining how effective the checks were at screening travellers and estimating at how disruptive it would be for those passengers and the operation of the airport.
Complex environment with significant interactions
An airport terminal is a complex space where tens of thousands of people can pass through each day, doing many different activities as they do so, including: checking in, shopping, eating and going through to the main security to catch their flights.
The security measures to be looked at were varied – including: step-through metal detectors, explosive sensors, sniffer dogs and camera-based standoff sensors – and in different locations, both at the entrance passages (or bridges) and the departure check-in concourse.
And the interaction between the people and the security measures is very important – queues could build, dogs may not be able to patrol effectively very crowded spaces, and the view of suspicious people for the standoff sensors could be blocked by other travellers.
Simulation modelling trialled on Heathrow T5 and 30,000 passengers
We realised that standard techniques used for modelling processes would not work for this problem, as they cannot handle these complex interactions. We believed agent-based modelling could meet all of these challenges and provide the types of answers the government wanted. We developed a proof-of-principle tool that focused on Heathrow Terminal 5’s departure concourse, modelling around 30,000 people that go through it on a typical day.
Modelling of transport arrivals, passengers, security personnel, sniffer dogs and flight departures
We developed a model of travellers arriving at the terminal building – by train, London Underground, bus, taxi and car – before checking in and going through the main security, with some of them going to shops and cafes first. We considered different combinations and configurations of the security measures, tracking every single passenger to see whether they were screened at least once and how long they took from arriving at the building to reaching passport control. By comparing the time it took the same passengers without the screening measures, we could calculate how much longer the passengers took. This showed how we could measure the impact and effectiveness of the security screening and how it could be optimised.
Wider application including helping the country in this time of crisis
Our model has clear potential for informing security operations at airports and other transportation locations. As we’re modelling every individual moving through the environment, interacting with others and systems while pursuing their particular goal, there is great opportunity to consider other applications. Of particular importance now is how the likes of airports, stations and shops can be configured and operated while maintaining social distancing. Our tool could be reconfigured do that and help get these working as efficiently and safely as possible, for the benefit of customer, staff and the businesses themselves