EUSAS recently organised a successful online European Conference on the topic ‘Artificial intelligence in fire detection and security – without the hype’. It showed once again the importance of technological development for an industry endeavoured to protect lives with particular relevance to the fire and security industry.

The conference opened with a discussion on what Artificial Intelligence (AI) is. The general concepts as well as the history and starting blocks of AI were discussed. Also, the current application fields for AI in Smart Living, as well as important requirements for the realisation of intelligent Smart Living services, were presented. Focal point for the fire industry were the presentations during the session on the benefits and opportunities of AI for fire detection and security.

Artificial Intelligence and Fire Safety

In his presentation, Guillermo Rein of the Imperial College London presented an innovative fire protection system that combines building sensors, computer modelling, and artificial intelligence (AI). It is called The Fire Navigator and aims to forecast the movement of a fire inside a large building, providing the fire brigades with essential information about flames and smoke ahead of time. It bridges the gap between fire safety and Building Information Modelling by making use of the data already produced by high-rise building sensors such as smoke and heat sensors.

A fast and simple cellular automata model assimilates sensor data, and via inverse modelling and genetic algorithm techniques, it uncovers the ignition location, time, flame spread rate and smoke velocity. A test case with synthetic data was shown for a real iconic building in London. The Fire Navigator concept would be specially suited for the protection of higher-risk buildings like high-rise and hospital, or key infrastructure like tunnels and power plants.

Sensor technology

Paul van der Zanden, General Director of Euralarm, elaborated on the connection between AI and the fire industry. He took a holistic approach by defining AI as ‘’Technology used to add value and/or improve the outcome of an existing or new process/system’’. The fire industry has a wide scope and covers many aspects. Within Euralarm fire safety is seen as an ecosystem and therefore fire safety should be part of the development process.

The question is if one can use other future spin-off developments from the AI world for the fire safety world

Assuming that everything is done in the design to prevent a fire from starting there still is a chance that a fire incident will happen. A key factor that defines the impact of this incident is time. Timely detection and sensitivity for unnecessary alarms have a relation with each other. Both factors can be improved by using new technologies including AI technologies.

The question is if one can use other future spin-off developments from the AI world for the fire safety world. The introduction of new sensor technologies available could be one of these spin-offs. With an example from AI sensor technology development, Paul van der Zanden showed how future fire detection can be brought to the next level.

AI-related vulnerabilities

In his presentation, Ibrahim Daoudi of CNPP presented the vulnerabilities related to the use of artificial intelligence on security/safety products. There are mainly 3 categories of vulnerabilities. The first category consists of adversarial attacks where the aim is to generate data sufficiently modified to mislead the model. The second category concerns physical attacks.

It is in fact based on adversarial attacks but applied to real objects. The third category is the traditional attacks on information systems leading to the poisoning of the model itself or its training data. All three vulnerability categories were discussed and explained.

Temporal Deep Learning

A new way to detect and localise smoke within such sequences was presented, called cell-wise classification

Utilising temporal information is crucial to detect smoke in video sequences. In his presentation, Andreas Wellhausen of Bosch Sicherheitssysteme GmbH presented the work on temporal approaches based on Deep Learning that are applied to Video Smoke Detection. Two methods were elaborated. Firstly, a combination of convolutional neural networks (CNN) and long-short-term-memory networks (LSTM), secondly the inflated 3D architecture (i3D), which consists of 3D convolutions.

These are two state-of-the-art approaches to extract spatial and temporal information out of video sequences. A new way to detect and localise smoke within such sequences was presented, called cell-wise classification. Furthermore, the advantage of temporal approaches over CNN methods, which are commonly used for detection problems in Computer Vision, was shown.

Training AI on synthetic data

Philip Dietrich of Bosch Sicherheitssysteme analysed the idea of using synthetic data to train Deep Learning Systems for Video-Based Smoke Detection algorithms. Compared to real data, gathering a large-scale database is significantly easier for synthetic data. It was shown how Deep Learning networks can be trained on synthetic videos. The results were compared with real data.

Experimental results support the hypothesis, that domain adaptation improves the generalisation of real data

As a means of bridging the domain gap between real and synthetic data, the concept of domain adaptation will be introduced. By forcing networks to extract similar features from real and synthetic data respectively, potential artifacts in synthetic data may not be learned by the network. Experimental results support the hypothesis, that domain adaptation improves the generalisation of real data.

Legislation and outlook

While the rapid adoption of AI creates exciting new opportunities for industry and individuals alike, it also poses an important question: does current laws apply to AI? Tadas Tumėnas of Orgalim discussed if and how this new technology should be regulated. He outlined the state of play of AI in Europe. He focused on the definition of AI which should be the essence of the EU legislative framework and presented the Commission’s work related to AI.

In the last presentation Lance Rütimann, chair of the Fire Section of Euralarm, said that if the fire safety industry does not take on the task of working with legislators, regulators, and standardisation bodies in defining the aforementioned regulatory landscape, then someone else will.

This is because the use of Artificial Intelligence to protect lives and assets makes good sense. Understandably, the path ahead is not clear, and there are many, many questions. The fact that the results of the work of the fire safety industry make the world a safer place for millions of people is the best motivation to set the focus on a new horizon.

Share with LinkedIn Share with Twitter Share with Facebook Share with Facebook
Download PDF version Download PDF version

In case you missed it

Panasonic AI-driven cameras empower an expanding vision of new uses
Panasonic AI-driven cameras empower an expanding vision of new uses

Imagine a world where video cameras are not just watching and reporting for security, but have an even wider positive impact on our lives. Imagine that cameras control street and building lights, as people come and go, that traffic jams are predicted and vehicles are automatically rerouted, and more tills are opened, just before a queue starts to form. Cameras with AI capabilities Cameras in stores can show us how we might look in the latest outfit as we browse. That’s the vision from Panasonic about current and future uses for their cameras that provide artificial intelligence (AI) capabilities at the edge. Panasonic feels that these types of intelligent camera applications are also the basis for automation and introduction of Industry 4.0, in which processes are automated, monitored and controlled by AI-driven systems. 4K network security cameras The company’s i-PRO AI-capable camera line can install and run up to three AI-driven video analytic applications Panasonic’s 4K network security cameras have built-in AI capabilities suitable for this next generation of intelligent applications in business and society. The company’s i-PRO AI-capable camera line can install and run up to three AI-driven video analytic applications. The AI engine is directly embedded into the camera, thus reducing costs and Panasonic’s image quality ensures the accuracy of the analytics outcome. FacePRO facial recognition technology Panasonic began advancing AI technology on the server side with FacePRO, the in-house facial recognition application, which uses AI deep learning capabilities. Moving ahead, they transitioned their knowledge of AI from the server side to the edge, introducing i-PRO security cameras with built-in AI capabilities last summer, alongside their own in-house analytics. Moreover, in line with the Panasonic approach to focus more on collaboration with specialist AI software developers, a partnership with Italian software company, A.I. Tech followed in September, with a range of intelligent applications, partially based on deep learning. Additional collaborations are already in place with more than 10 other developers, across the European Union, working on more future applications. i-PRO AI-capable security cameras Open systems are an important part of Panasonic’s current approach. The company’s i-PRO AI-capable cameras are an open platform and designed for third-party application development, therefore, applications can be built or tailored to the needs of an individual customer. Panasonic use to be a company that developed everything in-house, including all the analytics and applications. “However, now we have turned around our strategy by making our i-PRO security cameras open to integrate applications and analytics from third-party companies,” says Gerard Figols, Head of Security Solutions at Panasonic Business Europe. Flexible and adapting to specific customer needs This new approach allows the company to be more flexible and adaptable to customers’ needs. “At the same time, we can be quicker and much more tailored to the market trend,” said Gerard Figols. He adds, “For example, in the retail space, enabling retailers to enhance the customer experience, in smart cities for traffic monitoring and smart parking, and by event organisers and transport hubs to monitor and ensure safety.” Edge-based analytics offer multiple benefits over server-based systems Edge-based analytics Edge-based analytics offer multiple benefits over server-based systems. On one hand, there are monetary benefits - a cost reduction results from the decreased amount of more powerful hardware required on the server side to process the data, on top of reduction in the infrastructure costs, as not all the full video stream needs to be sent for analysis, we can work solely with the metadata. On the other hand, there are also advantages of flexibility, as well as reliability. Each camera can have its own individual analytic setup and in case of any issue on the communication or server side, the camera can keep running the analysis at the edge, thereby making sure the CCTV system is still fully operational. Most importantly, systems can keep the same high level of accuracy. Explosion of AI camera applications We can compare the explosion of AI camera applications to the way we experienced it for smartphone applications" “We can compare the explosion of AI camera applications to the way we experienced it for smartphone applications,” said Gerard Figols, adding “However, it doesn’t mean the hardware is not important anymore, as I believe it’s more important than ever. Working with poor picture quality or if the hardware is not reliable, and works 24/7, software cannot run or deliver the outcome it has been designed for.” As hardware specialists, Figols believes that Panasonic seeks to focus on what they do best - Building long-lasting, open network cameras, which are capable of capturing the highest quality images that are required for the latest AI applications, while software developers can concentrate on bringing specialist applications to the market. Same as for smartphones, AI applications will proliferate based on market demand and succeed or fail, based on the value that they deliver. Facial recognition, privacy protection and cross line technologies Panasonic has been in the forefront in developing essential AI applications for CCTV, such as facial recognition, privacy protection and cross line. However, with the market developing so rapidly and the potential applications of AI-driven camera systems being so varied and widespread, Panasonic quickly realised that the future of their network cameras was going to be in open systems, which allow specialist developers and their customers to use their sector expertise to develop their own applications for specific vertical market applications, while using i-PRO hardware. Metadata for detection and recognition Regarding privacy, consider that the use of AI in cameras is about generating metadata for the detection and recognition of patterns, rather than identifying individual identities. “However, there are legitimate privacy concerns, but I firmly believe that attitudes will change quickly when people see the incredible benefits that this technology can deliver,” said Gerard Figols, adding “I hope that we will be able to redefine our view of cameras and AI, not just as insurance, but as life advancing and enhancing.” i-PRO AI Privacy Guard One of the AI applications that Panasonic developed was i-PRO AI Privacy Guard Seeking to understand and appreciate privacy concerns, one of the AI applications that Panasonic developed was i-PRO AI Privacy Guard that generates data without capturing individual identities, following European privacy regulations that are among the strictest in the world. Gerard Fogils said, “The combination of artificial intelligence and the latest generation open camera technology will change the world’s perceptions from Big Brother to Big Benefits. New applications will emerge as the existing generation of cameras is updated to the new open and intelligent next generation devices, and the existing role of the security camera will also continue.” Future scope of AI and cameras He adds, “Not just relying on the security cameras for evidence when things have gone wrong, end users will increasingly be able to use AI and the cameras with much higher accuracy to prevent false alarms and in a proactive way to prevent incidents." Gerard Fogils concludes, “That could be monitoring and alerting when health and safety guidelines are being breached or spotting and flagging patterns of suspicious behaviour before incidents occur.”

What is the best lesson you ever learned from an end user?
What is the best lesson you ever learned from an end user?

Serving customer needs is the goal of most commerce in the physical security market. Understanding those needs requires communication and nuance, and there are sometimes surprises along the way. But in every surprising revelation – and in every customer interaction – there is opportunity to learn something valuable that can help to serve the next customer’s needs more effectively. We asked this week’s Expert Panel Roundtable: what was the best lesson you ever learned from a security end user customer?

What is the impact of remote working on security?
What is the impact of remote working on security?

During the coronavirus lockdown, employees worked from home in record numbers. But the growing trend came with a new set of security challenges. We asked this week’s Expert Panel Roundtable: What is the impact of the transition to remote working/home offices on the security market?