A few years ago, I was invited to talk about security and crowd safety to the board of directors at a major oil company headquarters in Aberdeen, Scotland. Since this was my home town, I did not think I would have any security problems when checking into the building.

However, at reception, I had to produce two photographic ID references (passport and drivers licence), was checked with a wand (metal detector scan), patted down and there was a thorough search of my computer bag before being issued with a radio frequency ID badge that would open certain doors and track me through the building. Finally, I was escorted to the board room to begin my presentation.

“I think you will agree, Professor Still, this is probably the safest building you have ever been in,” said the managing director, in a smug tone.

He did not understand the difference between security and safety. “Suppose I wanted to destroy your business?” I said. “Would I need to enter the building?” I added, turning over the security badge to reveal ‘in case of an emergency please assemble in the main car park’, indicated with a neat little diagram of the assembly point. By coincidence, I had parked my car next to the assembly point. “No! I would phone in a bomb threat and you would evacuate the building and assemble here, right next to my parked car with the bomb. Bang! I would take out your staff, your board members, and destroy your business without going near the building.”

He looked at me, his jaw opened, and the colour left his face.

That, in a nutshell, is the problem security is facing now – hardening the building security and overlooking the soft targets, such as the entry and assembly points in the system. As an example, security at airports may be increased and, as a direct result, crowds will gather in easily accessible areas.

There are many such soft targets around the world that have already been attacked: airports (Brussels and Glasgow), stadia (Paris), sporting events (Boston Marathon), theatres (Paris) and transport systems (London); in fact, anywhere a crowd gathers is a potential target. Increasing access control/screening/security will often lead to larger crowds gathering in unsecured areas.

So how is security and safety balanced? How are the potential soft targets and the queueing crowds reduced? In the example above, the company thought that the business was the building. So, by securing the building, the directors thought they had secured the company’s future, but they overlooked the fact that the the business cannot function without its staff.

There are two elements to the threat – getting people into secure areas as quickly as possible and evacuating people to safe areas. Using an airport as an example again, increased security, specifically the airport screening process, results in a large number of people in a very confined space.

The first part of the problem relates to the process called ‘queueing theory’, which is the analysis of the security screening process and the size of the queueing crowd. In essence, the security screening process is similar to a supermarket checkout system.

Consider the situation for both systems. There is an arrival rate (the rate at which people arrive at the screening system), and a service rate (the rate at which items are checked). The objective is to minimise the queue and process all the items in the shortest time, without missing anything. The mathematics for the supermarket checkout system uses the same formulae as the security screening process. The overall system demand (most popular times for arrival) can be evaluated and the system manned to minimise the queues. The infrastructure is also similar; both systems use devices that scan items.

Does Increasing Security Provide a Safer Environment?

Well, it does feel safer once a person (eventually) passes the screening process, but increasing security screening does not reduce the potential threat to the crowd in the pre-screening areas, as a recent BBC headline indicated.


The mathematics of the supermarket checkout reveals a hidden problem; the same problem that is being observed in the security screening process. A contingency must always be provisioned in the system for the odd item that causes a delay.

The dramatic increase in the numbers of people queueing due to a slight change in the screening process needs to be understood to balance security and safety, and to reduce the creation of soft targets. I recently passed through a security screening process that took 45 minutes; previous trips at the same airport took less than 10 minutes. The bottleneck was not the machine; it was the number of staff processing items that the machine rejected. Add one more person to the human processing/checking of items and the waiting time could have been significantly reduced.

Over the last few decades of applying the principles and applications of crowd dynamics (the mathematics and psychology of crowd analysis), queueing applications have proven to be simple to explain. Any increase in the screening process will result in a much larger increase in the queueing process. The supermarket systems balance demand with the flexible checking resources (more checkouts at busier time). When equipment and space are at a premium, for example in the airport, then the efficiency of the system is entirely dependent on human factors. Bottleneck analysis and activity cycle analysis (cycle time analysis) are not complex, but do require an application of mathematics, specifically, queueing theory.

Evacuating Crowds to Safe Spaces

As the oil company example at the beginning of this article highlighted, the threat may be the assembly area. These were defined for fire scenarios, but the use of primary and secondary devices, where the crowd is driven towards the secondary device, is a frightening tactic. During the height of the Irish Republican Army (IRA) bombing in the UK, a stay-put approach was developed – keeping people inside buildings proved to be a safer alternative to evacuation.

Organisations often create documents and procedures that consider a wide range of threats with a wide range of responses. This can lead to a complex and difficult to apply prescriptive set of conditions, with uncertainty if the threat is not in the list. A tactical approach would be to consider the options for safe evacuation. There are five possible crowd dynamics to consider:

  1. total evacuation: all exits are viable, everyone out as quickly as possible
  2. directed evacuation: clearing the people away from the area of the threat
  3. phased evacuation: decanting an area section by section
  4. stay-put: the building/infrastructure offers an element of protection
  5. invacuation: move everybody out of sight, from open spaces into buildings

When considering the above, regardless of the nature or location of the threat, the key to a successful implementation is the communication process with the affected individuals – the crowd. For example, a fire bell can alert in the first instance of the above dynamics (total evacuation), but does not communicate direction for the other scenarios.

Consider the environment, how the location of the threat would be communicated to the site and the required crowd dynamic. If occupants need to be informed to leave the site via a specific route, how are they informed? Sometimes leading people away from a threat (follow me) can be more effective than pushing people (get out). If the problem is approached from the required action of the crowd (total, directed, phased, stay-put or invacuation), it can lead to a much clearer, simpler and easier to follow contingency plan.

In summary, increasing security can lead to increasing the queueing crowds and creating soft targets. A balance between security and safety must be assessed, understood and applied. Understanding the process of communicating with the crowd, especially during an evacuation, is essential to both security and safety. Do not underestimate the importance of these two elements of crowd dynamics.

Keith Still
Dr G. Keith Still is a Professor of Crowd Science at Manchester Metropolitan University and Director of Crowd Risk Analysis (UK). He is a crowd risk analysis specialist who has lectured at the UK Cabinet Office Emergency Planning College, contributing to their crowd safety-related courses.