Why You Should Read This Introduction and What to Expect

Reading and comprehending a technical text represents a substantial investment of time and energy on the part of the reader. Digital security systems are well-known in the industry and have been applied for decades, beginning with Closed Circuit Television (CCTV) and Video Recorders. What is driving the need for a new entry in the technical literature for Digital Security and Safety Systems?

Two megatrends are driving a revolution in the purpose and design of Digital Security Systems: the assets being protected and the rise of Artificial Intelligence (AI) as a pragmatic technology.

…Because That’s Where the Money Is

Traditionally, the assets being protected consisted of physical items that could be touched – currency, gold, bearer bonds. When Willy Sutton, an infamous American bank robber, was asked why he robbed banks, his apocryphal reply was simply – “Because that’s where the money is” (Though Mr. Sutton denied ever saying it). During his career from the 1920s to the 1930s, it is estimated Mr. Sutton stole $2M. Allowing for a bit of inflation, Table 1-1 shows the characteristics of $10M in assets of various forms.

Table 1-1 $10M in Assets
Full size table

The physical assets lend themselves well to the traditional methods of guns, guards, and gates as all require physical transport to be stolen. In all but the case of gems, the physical characteristics for even a modest sum of $10M require transport of hundreds of pounds of material and substantial physical bulk. However, an increasing proportion of the world’s assets are embodied in digital form as digital data. These digital assets take many forms – trade secrets, electronic currencies, business data related to customers, processes, and methods. These digital assets require no physical access to steal or corrupt. Guns, guards, and gates are of limited use in protecting these assets. Consequently, there is a fundamental shift in the types of threat models emerging to attack these new digital assets.

Cogito Ergo (Multiply and) Sum – Artificial Intelligence

The second megatrend impacting the Digital Security and Safety industry is the rise and widespread adoption of Artificial Intelligence (AI) techniques and methods for practical applications. AI experienced several cycles of enthusiasm and disappointment starting in the mid-twentieth century. In the early 2010s, the introduction of Convolutional Neural Networks combined with the evolution of the requisite compute power enabled practical demonstrations of computer vision-based AI. The subsequent decade saw the continued co-evolution of AI algorithms and specialized compute platforms. The immediate result was the ability to capture, analyze, and act upon vast quantities of data in real-time. Security systems progressed from being reactive to becoming proactive, real-time capabilities. Many of the functions that once required human observation and reaction are now being delegated to machines. So, while new and novel threat models are emerging, there are also new and novel techniques based on AI to counteract and mitigate these new threat models.

What This Means for You As a Security and Safety Professional

The world is changing rapidly – not only the assets being protected but also the tools and techniques used to protect both traditional and emerging assets. The purpose of this text is to assist security and Safety professionals in understanding and addressing these contemporary trends and techniques, to build on the skills and knowledge you already possess to extend to these new regimes. Our goal is to help you understand how to navigate this new landscape, introduce basic concepts and techniques, and enable you to collaborate with and incorporate these new techniques into your professional life.

It is also important to be clear about what is beyond the scope of this text. This is not intended to be a detailed investigation and exploration of the mathematical foundations of AI and Machine Learning (ML). There are a number of academic texts which cover these areas quite adequately. It is beyond the scope of this text to enable the reader to create novel AI algorithms or discuss in detail the current state-of-the-art algorithms. The reason is quite simply that the field is in a very fast state of evolution, often requiring less than six months from the creation of a new AI algorithm to productization and within a year or two – obsolescence. By the time this tome is published, much of such an attempt would be outdated. For the detailed current state of the art, we will defer to specialized conferences and journals on the topic. Fortunately, the vast majority of the intended readership will require the ability to make use of the outputs of the research and development, and those wishing to pursue an in-depth study will still find this a useful introduction.

What we will enable you to do, however, is intelligently converse with and collaborate with specialists in the field of AI-enabled Security and Safety technologies. Our intended audience includes, but is not limited to:

  • Security executives responsible for protecting corporate and governmental assets

  • Security system specifiers creating system Requests For Proposals (RFP)

  • Security system architects are responsible for creating detailed systems architecture requirements

  • Security system consultants guide clients in best practices and proposal evaluation

  • Security system designers must ensure their systems protect the right assets with modern techniques

  • Security system installers and Value Added Resellers who need to understand the systems being installed and verify the systems perform correctly

  • End users who need to understand the security systems they have purchased and what those systems will and will not protect in terms of assets and threats

Every Journey Begins with a Single Step – Maya Angelou

If you have read this far, then you understand that the security world is evolving and you are ready to begin that journey with a single step, the first of the journey. This describes some of the main features and points of the journey we will take together.

The author’s goal is to inform you about how intelligent multimodal security systems (IMSS) can help you sense the world to make the world more secure, especially the security you need in the system, to better assure that outcome. Note that while following these recommendations will make your systems more secure, no component, device, or system can be absolutely secure.

Chapter two is a system-level view of Intelligent Multimodal Security Systems. You will understand the history, current state, and trends of these systems. The basic system components are described, and you will get an introduction to the security of security from the devices at the sensor edge to the cloud.

In Chapter three, you will dive deeper into the evolution of security systems and learn what an Intelligent Multimodal Security System does and how it is done.

In Chapter four, you will gain detailed understanding of security analysis methods with a focus on the threats and assets for security systems, followed by the step-by-step elements necessary to build a secure system.

Chapter five gives you the detailed knowledge of security for the wave of the future: The intelligent aspect of Security Systems. Machine learning has a number of unique risks and challenges that must be understood and addressed for a security system that includes artificial intelligence or machine learning capabilities. You will learn about the unique assets and threats to machine learning, how to provide protection for the applications and the data, and ultimately, trust for this class of security systems.

In Chapter six, you will see these principles in practice through three system examples: a small business system, and edge server system, and a large system in a smart city. Through these examples, you will see the kinds of threats in these types of systems and how to mitigate them.

Devices and networked processing and content aggregation systems are evolving quickly. Laws and regulations and corresponding standards that impact these systems are another rapidly evolving element you need to stay abreast of. Perhaps the only things evolving faster are the attacks on those systems. Chapter 7 will guide you on ways to keep up with the change in the systems and how to defend those systems.

Indeed, there is so much change that we saved the last updates for a final “as we go to press” refresher in Chapter 8, so you will get the most recent updates we can provide.