Written by Miltiadis “Miltos” Alamaniotis
The smart cities vision entails the continuous collection of data and subsequent utilization of information that is communicated in every possible direction. Its implementation requires the extensive use of sensors, information and decision-making technologies. The development and deployment of smart city technologies will benefit several aspects of residents’ daily lives. Amongst them, security is a preeminent component that will only be enhanced in smart cities. A crucial component of the overall city security architecture is that of nuclear security, which refers to the detection, identification, and localization of nuclear materials that may be used in terrorist activities. This article explores the potency of data analytics as a means for implementing smart nuclear security within the vision of smart cities.
The terror attacks that took place in September of 2001 (known as 9/11) have made authorities revisit overall security architecture. Specifically, new security vulnerabilities were identified and the potential threats from those were fully examined . One of the scenarios that was identified after 9/11 was the possibility of an attack using nuclear and radioactive materials . This type of attack would have lethal consequences and its severity would be much higher than any other conventional attack. Detonation of a device containing nuclear material within a metropolitan area is an unwanted scenario, especially given that the material's dispersion would remain in the environment for a long time after the attack.
A significant line of defense against this type of terrorism includes the detection, identification, and localization of nuclear material before its use (explosion), known by the term nuclear security . Up to this point, nuclear security utilized radiation measurement systems to analyze the obtained data, and identify patterns associated with nuclear material activities. Though radiation systems and their data analysis modules are widely used, they have several limitations that may be exploited by potential terrorists. For instance, radioactive material may be shielded (e.g., by lead) making their detection highly challenging.
Data Analytics Enabling the Vision of Smart Nuclear Security in Smart Cities
Federated Learning is called "privacy by design." As TAI also considers privacy Smart city technologies provide new technologies for implementing smart nuclear security. In particular, smart city technologies may provide new signatures and observables that are associated with the storage, movement, and use of nuclear materials in nefarious activities . Signatures refer to pieces of information that are uniquely associated with specific nuclear materials, while observables refer to the observed features of a signal that may lead to the identification of the signatures.
Data analytics consists of the bridge between smart cities and nuclear security technologies as presented in Figure 1. In this article, smart nuclear security is defined as the utilization of data analytics to prevent the use of nuclear material in terroristic activities within a city’s premises. Figure 2 highlights those smart city components that are envisioned to contribute to enhancing the overall nuclear security in metropolitan areas with the use of data analytics.
Figure 1: Visual representation of data analytics for implementing nuclear security in smart cities.
Figure 2: Smart city components and their relation to nuclear security.
More particularly, the vision contains:
Dynamic sensor networks: nuclear security will greatly benefit from the use of smart transportation infrastructure. Interconnected vehicles may consist of a line of defense against the movement of nuclear materials. In particular, vehicles may be equipped with radiation sensors that obtain consecutive measurements, which are transmitted through the data network and are processed to make inferences. Sensor instalment in vehicles is a very convenient choice given that it may create a sensor network with a dynamic architecture that has no specific form.
This dynamically varying architecture of the sensor network has the following benefits:
- Does not allow terrorists to have prior knowledge of sensor locations.
- Allows the fusion of data from various sensors within close proximity.
- Greatly increases the probability that a vehicle (sensor) will pass close to nuclear material that is in motion. This detection may be done using a single sensor or with multiple sensors .
- Detection of a source may be verified by driving sensors close to the area of detection for verification.
At this point, it should be noted that the vision of smart nuclear security encompasses all the vehicles within the city premises being part of the sensor network. However, in a realistic scenario, the network is mainly comprised of public transportation means since they move within the city limits for long periods of time during the day.
Smart Buildings as Radiation Sensors
A smart building may also be part of the sensor networks. Sensors may be installed at key points and take radiation measurements at its entrance points and in the periphery of the building. Sensors in buildings may also be used together with the moving sensors from vehicles, thus creating a powerful network that covers the city’s whole area. With this type of network, illicit movement of nuclear material within the city will likely be picked up by at least one sensor: a terrorist carrying a nuclear device will certainly pass by a building or public transportation.
Smart Energy Networks as Sensors
The electrical energy grid may also be considered to be a sensor network. Notably, the power grid may be seen as a ubiquitous network given that it may be found everywhere in cities. In other words, distribution lines and power grid components are found in almost every inch of a city, hence making the power grid a potential sensor network. The electromagnetic interferences caused by nuclear material close to distribution lines and other power grid equipment may provide new observables for the illegal movement of nuclear material.
In that case, detection and identification of electromagnetic interferences would be performed via advanced data analytics. Supervisory control and data acquisition (SCADA) and phase measurement unit (PMU) devices may also be used for collecting and analyzing data and making inferences related to illicit nuclear activities.
Digital Citizens as Sensors
TIt is expected that one of the basic components of smart cities is the digital form of the citizens. In other words, the citizens will be able to connect to the city’s infrastructure through digital devices like smartphones and tablets and obtain useful information.
Here, we envision those citizens will also be able to provide information to the infrastructure for the purposes of nuclear security. For instance, citizens who notice suspicious behavior may report it to the city, and they may also report a peculiar observation (e.g., a citizen carrying lead bricks). In addition, advancements in nanotechnology may allow citizens to carry radiation detectors as part of their phones. To make it clear, the miniaturization of radiation sensors will allow their incorporation into smartphones. The ubiquitous presence of smartphones in cities accommodates full coverage of the city against nuclear threats.
In the above scenarios, data analytics would be used for:
- detecting patterns of interest in data obtained by detectors in smartphones
- identifying false behaviors in reported suspicious activities
- fusing data coming from a variety of citizens
In this article, a vision for implementing smart nuclear security in the context of smart cities is given. The extensive use of data analytics is essential in the realization of this vision. Smart city components may provide opportunities for obtaining new signatures and observables via the use of data analytics by analyzing multidimensional information vectors. The challenge in this vision is to exploit every piece of information coming from the various and uncorrelated data sources within the city to contribute to the detection of illegal nuclear activities.
- A.V. Putilov, V. N. Chervyakov, and P.I. Kolykhalov. “‘Smart’ nuclear cities and training.” In IOP Conference Series: Earth and Environmental Science, vol. 740, no. 1, p. 012014. IOP Publishing, 2021.
- C. Talmadge, “Deterring a nuclear 9/11,” Washington Quarterly, vol. 30(2), 2007, pp. 21-34.
- Alamaniotis, M., “Fuzzy Integration of kernel-based Gaussian Processes applied to Anomaly Detection in Nuclear Security,” 12th International Conference on Information, Systems and Applications (IISA 2021), July 2021, pp. 8.
- M. Alamaniotis, and A. Heifetz, A., “Survey of Machine Learning Methodologies in Radiation Data Analytics pertained to Nuclear Security,” Advances in Machine/Deep Learning-based Technologies, Book Edited by. G. Tsihrintzis, M. Virvou, L. Tsoukalas, A. Esposito and L. Jain, vol. 23, Springer, 2022, Chapter 6, pp. 97-115.
- M. Alamaniotis, “Smart Data Analytic based Transform of Low-Count Gamma-Ray Spectra for Enhancing Isotope Detection by Using a Self-Learning Window of Relevance Vector Regression,” IEEE Transactions on Nuclear Science, vol. 69(6), June 2022, pp. 1357-1365.
This article was edited by Qi Lai.