Nowadays the work of a surveillance officer is complex. It is expected of them to respond proactively to what happens around them in order to prevent incidents form happening instead of responding to them. More and more information has to be dealt with; besides the growth of the amount of cameras, the increase of other types of sensors that generate data and information is consistent. How should a surveillance officer deal with these changes?
Surveillance officers observe people, they select people on who they focus a little longer, and in some cases they undertake action. By asking them a few simple questions for example. For the most part this process takes place based on intuition. The effectiveness of (proactive) surveillance benefits from more knowledge on human behavior, which will result in the fact that surveillance officers can look at their work environment in a more objective manner. This will also make them less vulnerable to cognitive bias such as stereotyping and tunnel vision. In our projects, we do research on what type of behavior is relevant for the surveillance officers, how they can respond to these behaviors in an appropriate manner, how they can minimize cognitive bias, and how they can be supported by technology.
Worldwide, the amount of cameras used for security still grows significantly. The increasing amount of cameras is not always similar to the increase of the amount of operators behind the screens. How do we enlarge our detection abilities anyway? We focus on different areas that contribute to this capability. For example, not everyone has the same qualities when it comes to recognizing and responding to incidents. We study what competences are associated with good performance, which enables surveillance centers to select, recruit and train in a more focused manner. We also develop intelligent software that supports operators and surveillance officers with proactive surveillance. For example by developing software that automatically analyses specific behavior or incidents (e.g. aggression, violence, left luggage). By doing this, relevant events are filtered from the enormous amount of information that operators have to deal with.
Cameras are not the only sensors that are used for security purposes. Another way to increase our detection abilities is to collect information or data with the use of other sensors (sound- or heatsensors, or even databases from the internet). In a couple of cities in the USA surveillance officers are working with statistical models that can predict (to a certain amount) where crime is going to take place. The first experiences are very positive and the results don't lie: in Santa Cruz burglary cases decreased with 30%. The Netherlands is also experimenting with this method. What kind of data do we need for this? What else can we do with this type of knowledge? Ethical, societal and juridical questions are at the foundation of this kind of research, so they also need to be taken into account before implementing these type of methods. We can help to develop this knowledge, and we can help to make it applicable.