Periodic Reporting for period 2 - EU-SENSE (European Sensor System for CBRN Applications)
Période du rapport: 2019-11-01 au 2021-10-31
The EU-SENSE project put forward the following three objectives:
• High-Level Objective 1 – To contribute to better situational awareness of the CBRNe practitioners through the development of a novel network of chemical sensors, which will provide a technological solution to relevant gaps presented in the ENCIRLCE catalogue of technologies.
• High-Level Objective 2 – To improve the detection capabilities of the novel network of chemical sensors through the use of machine learning algorithms to reduce the impact of environmental noise and the application of contaminant dispersion models.
• High-Level Objective 3 – To showcase the usability of the EU-SENSE network to CBRNe practitioners in order to validate the system and to maximize its exploitation potential. The objective also entails the preparation of training sessions with CBRNe practitioners in relevant conditions.
These objectives are being achieved through the below-presented science and technology objectives:
• S&T Objective 1: To improve the detection of large-spectrum of chemical agents via a novel network of sensors
• S&T Objective 2: To improve detection accuracy and minimize false alarm rate
• S&T Objective 3: To provide novel capabilities for the training of CBRNe practitioners
An important asset of the EU-SENSE, which falls into the category of societal impact, will be the improvement of citizens’ safety. The EU-SENSE will be used to prevent and respond to a potential chemical threat. As a consequence, the system will help to decrease detrimental effects to human health and/or the number of casualties. It is believed that the development of novel systems for chemical threat detection supported with relevant training tools will help CBRNe practitioners perform their daily duties more effectively and provide a high detection coverage of the target area during, e.g. mass event. Complex detection systems, such as EU-SENSE, will offer a chance for better operation in response mode assisting practitioners at the estimation of contamination source and further hazard prediction. What is more, the capabilities and operational skills of the practitioners will be increased thanks to the training mode, which will allow users to simulate potential threats and practice a range of countermeasures.
• High-Level Objective 1 – To contribute to better situational awareness of the CBRNe practitioners through the development of a novel network of chemical sensors, which will provide a technological solution to relevant gaps presented in the ENCIRLCE catalogue of technologies.
• High-Level Objective 2 – To improve the detection capabilities of the novel network of chemical sensors through the use of machine learning algorithms to reduce the impact of environmental noise and the application of contaminant dispersion models.
• High-Level Objective 3 – To showcase the usability of the EU-SENSE network to CBRNe practitioners in order to validate the system and to maximize its exploitation potential. The objective also entails the preparation of training sessions with CBRNe practitioners in relevant conditions.
These objectives are being achieved through the below-presented science and technology objectives:
• S&T Objective 1: To improve the detection of large-spectrum of chemical agents via a novel network of sensors
• S&T Objective 2: To improve detection accuracy and minimize false alarm rate
• S&T Objective 3: To provide novel capabilities for the training of CBRNe practitioners
An important asset of the EU-SENSE, which falls into the category of societal impact, will be the improvement of citizens’ safety. The EU-SENSE will be used to prevent and respond to a potential chemical threat. As a consequence, the system will help to decrease detrimental effects to human health and/or the number of casualties. It is believed that the development of novel systems for chemical threat detection supported with relevant training tools will help CBRNe practitioners perform their daily duties more effectively and provide a high detection coverage of the target area during, e.g. mass event. Complex detection systems, such as EU-SENSE, will offer a chance for better operation in response mode assisting practitioners at the estimation of contamination source and further hazard prediction. What is more, the capabilities and operational skills of the practitioners will be increased thanks to the training mode, which will allow users to simulate potential threats and practice a range of countermeasures.
The hereby section is to provide insight into the project progress towards the above set of project objectives. First of all, the progress towards the high-level objective 1 is marked by the advancements made in the sensor node development domain. Throughout the reporting period, the consortium investigated the sensor hardware, their communication protocol and interfaces in order to come up with a design of the sensor node. This process also required the implementation of several amendments in sensor hardware/software in order to reduce the occurring stability issues. The key achievement made in the period is the release of the 1st prototype of the sensor node. The purpose of this piece of hardware and software is the integration of the data coming from the applied sensors in the EU-SENSE system. The node integrates data from the following sensors: Proengin AP4C (Flame Photometric Detector), AIRSENSE Gas Detection Array – Personal (Ion-mobility spectroscopy and Photoionization technology), AIRSENSE GAS Detection Array – Personal (Ion-mobility spectroscopy and Electrochemical cell technology), and TNO SRD detector (based on Metal Oxide technology). Furthermore, it is essential to touch upon the aspect of situational awareness improvement. In the reporting period, the consortium has designed the situational awareness components including situational awareness tool, hazard prediction tool, and threat source estimation tool. The purpose and functionalities of these tools is described, in more depth, in deliverables D3.1 D3.2 and D3.3.
Regarding the second high-level objective, the progress on novel capabilities of detection is mostly reflected through the definition of the data fusion component, which has been organized into a pipeline of activities including classification, identification, and concentration estimation. Furthermore, there have been also developments in the area of environmental noise learning. It has been defined that in the preparedness phase, anomaly detection will be applied in order to monitor unusual situations. In response phase, however, the system will feature “normality detection”, which relies on the comparison of unaffected reference with other sensors in the response phase. Further developments in these aspects will be achieved in the second part of the project, especially once the data from long-term measurements are available for analysis.
The last high-level objective of the project plans the development of dedicated training components of the system as well as showcasing the system with the involvement of end-users. In this area, the consortium has made progress by defining the design of the training component. It has been agreed that the simulation mode will operate on synthetic data and will have access to a database with historical data, which is beneficial for end-users as they have the opportunity to analyze and learn from past events. The details on training components are available in WP3 deliverables including D3.1 D3.2 D3.3 and D3.5. Demonstration of the system, which is realized under WP8, has been confirmed to be held in Nowy Dwór Mazowiecki (Poland). The spot is a professional training centre of first responders in Poland.
Regarding the second high-level objective, the progress on novel capabilities of detection is mostly reflected through the definition of the data fusion component, which has been organized into a pipeline of activities including classification, identification, and concentration estimation. Furthermore, there have been also developments in the area of environmental noise learning. It has been defined that in the preparedness phase, anomaly detection will be applied in order to monitor unusual situations. In response phase, however, the system will feature “normality detection”, which relies on the comparison of unaffected reference with other sensors in the response phase. Further developments in these aspects will be achieved in the second part of the project, especially once the data from long-term measurements are available for analysis.
The last high-level objective of the project plans the development of dedicated training components of the system as well as showcasing the system with the involvement of end-users. In this area, the consortium has made progress by defining the design of the training component. It has been agreed that the simulation mode will operate on synthetic data and will have access to a database with historical data, which is beneficial for end-users as they have the opportunity to analyze and learn from past events. The details on training components are available in WP3 deliverables including D3.1 D3.2 D3.3 and D3.5. Demonstration of the system, which is realized under WP8, has been confirmed to be held in Nowy Dwór Mazowiecki (Poland). The spot is a professional training centre of first responders in Poland.
Exploitable assets: In the context of EU-SENSE, there have been the following assets identified:
a) EU-SENSE Sensor Node – a heterogeneous node, which will be developed within the project. The node will incorporate various sensors and will be able to communicate with the headquarters.
b) The network of chemical sensors – a network of EU-SENSE nodes, which will be utilized for measurement data collection and during the final demonstration.
c) Unified Data Model – a standardized description of the sensors network which will facilitate communication between the system components.
d) Environmental Noise Learning Tool – a software algorithm that will utilize machine learning to filter the environmental noise and, as a consequence, reduce false alarm rates.
e) Source Location Estimation Tool – a software component responsible for calculating chemical threat source location.
f) Hazard Prediction Tool – an algorithm based on contamination modeling that will estimate the most probable future dispersion of the contaminant.
g) Situational Awareness Tool – a user-access point that will collect the data from various components and integrate them into the situational view.
h) EU-SENSE training mode - the EU-SENSE system mode that will be dedicated for training purposes. It will utilize the other components and artificial data in order to provide a realistic simulation environment.
i) Methodology and guidelines on how to implement the proposed approaches that make use of the aforementioned capabilities.
a) EU-SENSE Sensor Node – a heterogeneous node, which will be developed within the project. The node will incorporate various sensors and will be able to communicate with the headquarters.
b) The network of chemical sensors – a network of EU-SENSE nodes, which will be utilized for measurement data collection and during the final demonstration.
c) Unified Data Model – a standardized description of the sensors network which will facilitate communication between the system components.
d) Environmental Noise Learning Tool – a software algorithm that will utilize machine learning to filter the environmental noise and, as a consequence, reduce false alarm rates.
e) Source Location Estimation Tool – a software component responsible for calculating chemical threat source location.
f) Hazard Prediction Tool – an algorithm based on contamination modeling that will estimate the most probable future dispersion of the contaminant.
g) Situational Awareness Tool – a user-access point that will collect the data from various components and integrate them into the situational view.
h) EU-SENSE training mode - the EU-SENSE system mode that will be dedicated for training purposes. It will utilize the other components and artificial data in order to provide a realistic simulation environment.
i) Methodology and guidelines on how to implement the proposed approaches that make use of the aforementioned capabilities.