Critical infrastructures (CIs) are facilities that are basic to societal function and well-being, such as power plants,communication centrals, transportation facilities, water supply, hospitals and political sites. Such CIs frequentlyserve large areas, so that even a relatively minor incident or attack can have major consequences to a largepopulation. CIs must therefore be safeguarded to ensure their function and to keep the citizens safe from potentialhealth hazards. An incident at a CI may have effects that are not limited by country borders, making the protectionof CI a challenge with a European dimension. The European Council has therefore adopted Directive2008/114/EC on Critical Infrastructure Protection to help the Member States to identify critical infrastructures andto improve their protection. Furthermore, following the 9/11 attacks in New York, the Madrid bombing in 2004and the London attacks in 2005, the EU also launched the European Programme for Critical InfrastructureProtection to improve the safety of critical infrastructures.
Advances in sensor technology and autonomous systems promise improved security solutions for CIs withsignificant savings in manual labour. The proposed P5 project deals with the problem of proactively detectingpotential illegal physical intrusions attempts. Specifically, the P5 system will monitor the outer perimeter of theCI area and give early warning in case a potentially threatening situation arises. Such a warning before anintrusion takes place would give valuable time to implement safety procedures at the infrastructure to minimizethe potential damages.
Current day systems are inadequate for the early warning task. Commercial systems are often based on a singlesensor modality and cannot provide 24/7 operability in all weather conditions. Lacking the ability to recogniseanimals and other irrelevant objects, such as blowing debris and trees moving in the wind, false alarms are amajor problem. Interpretation of behaviour and intent is yet a higher cognitive level missing in current systems.Instead, interpretation is left to human operators, who will have to screen all incoming sensor data to obtain asolution that is proactive and not just reactive, alarming for example when a fence is climbed. Manual screening isa monotonic task prone to errors due to fatigue or moments of low concentration or distraction.
Another problem of surveillance arises with the use of visual cameras directed towards public areas, which is inconflict with fundamental human rights such as the right of liberty and the right to privacy and data protection.Thus, user requirements on a proactive surveillance system do not only stem from the operators of the CI, butalso from the people being monitored. This matter requires very careful considerations to find the balancebetween the need of security and the privacy interests of the European population.
The P5 Project
By design, the envisioned systems will handle current as well as expected future requirements on privacy preservation and surveillance system legislation. The aim is to describe a sensor system that for different protection contexts can be tuned to balance between potent but potentially invasive surveillance techniques, and the personal integrity of the citizens. The sensor system should adhere to a publicly accepted principle of proportionality. To monitor the privacy aspects, P5 has a group of independent advisors with expertise in ethics and law.
The P5 project is strongly user driven and the goal of P5 is to demonstrate beyond state-of-the-art surveillance abilities at the OKG nuclear plant outside Oskarshamn, and at CAST outside London. Both demonstrations will display new functions that are aimed at enabling a more cost effective perimeter protection of critical infrastructure in relevant scenarios, and that are privacy preserving, practically useful, and commercially successful. P5 goals also include contributions to the technological and scientific community in the form of articles, workshops and conference presentations, and contributions to the ethical and legal community in the form of reports and workshops. To make sure the technical achievements are in line with end user needs, P5 has an advisory stake holder group with international representation from both critical infrastructure protection and surveillance providers.
Threats of particular interest are terrorists, saboteurs and thieves approaching from land (walking or in vehicles), water (boats), or air (mainly small aircraft, UAVs). The envisioned system will give automatic and reliable alarms to an operator that make the final threat assessment. The alarms are based on radar, thermal and visual cameras collaborating in a network where sensor fusion and anomaly detection support efficient and robust surveillance of large areas with limited sightlines. The work within P5 has a focus on large area surveillance outside the physical perimeter (fence) of critical buildings. The objective is to give an early warning, that is, the operator should be given the opportunity to react on the threat well before it reaches the perimeter of the inner area of the critical infrastructure. Therefore P5 has the focus on the early warning surveillance area, see figure below. If the perpetrator manages to reach the fence undetected, there may be little chance left to prevent unauthorized ingress and a security breach of magnitude may then follow in worst case.
The P5 perimeter surveillance system has the following three basic goals:
- G1. Privacy preservation
- G2. Early warning with low false-alarm rate
- G3. 24/7 surveillance of a bubble-shaped perimeter
The research leading to these results has received funding from the European Community's Seventh Framework Programme Security Theme (10) under grant agreement number 312784. P5 started August 1, 2013, and will go on for three years. Budget total is approximately 4,000 Euro, where the funding comes to a part from the European Community, to a part from the partners of P5.