Article by Johan Paulsson, CTO, Axis Communications
It only feels like five minutes since we published our 2021 technology trends post. The past year has flown by. I hope it’s been a positive one for all of you, and though we’re still experiencing an uncertain environment, you’ve remained healthy and safe.
As ever, I tend to start thinking about next year’s technology trends in the context of those which we predicted for this year. However, having published annual technology trends posts for a while now, I looked back through them all to prompt some thoughts.
What’s very clear is that the concept of ‘trust’ has appeared in most of the posts in one form or another. The context has changed slightly over the years, sometimes focused on trust in the use of data, trust that systems are secure, or the fundamental need to trust that organizations are doing the right thing. But it’s clear that matching the ongoing pace of technology innovation through an equal evolution in building trust in the use of technology is vital. We don’t think that 2022 will be any different.
Now more than ever there’s a healthy desire – from individuals, organizations, and legislators – to ensure that new technologies are being developed, manufactured, used and secured in ways that we can trust are positive. For Axis, of course, that means always working towards our vision of a smarter, safer world.
As we move into 2022, it’s interesting to see how many of the technology trends we see for year can be linked back to the need to build a trusted technology ecosystem.
Universally connected across hybrid environments
To the end user of technology – from a consumer using their mobile phone to security personnel managing video surveillance – the technology architecture being used to deliver services has become invisible. Whether processing takes place on a device, local server or in a remote data center doesn’t matter: everything is simply ‘connected’.
Last year we talked about the world going ‘horizontal’, where the combination of cloud, on-premise server and edge technologies would be increasingly used together, each employed to its strengths, in so-called hybrid solutions.
This hasn’t changed, but it’s very apparent that the question of architecture is unique to each customer and needs to take account of both internal resources and policies, and of external factors such as local and international regulation.
As a security solution vendor, it’s not up to us to define for the customer the environments and architecture they should use, but rather to equip them with the tools and flexibility to decide on the best solution for their unique situation.
Given that ‘connected’ has become the default, we do believe that most surveillance solutions will ultimately be hybrid – indeed, many already are – which in itself has implications.
Cybersecurity born from a healthy skepticism
We don’t always think of skepticism as a positive trait, but in relation to cybersecurity it can be a prudent one.
The billions of connections that now exist between devices, networks, and data centers have made the concept of securing a perimeter around any organization almost completely obsolete. The walls that might have previously existed have become permeable, and a new approach to security has therefore emerged: zero trust networks.
Regular readers of our annual technology trends post may question the move to zero trust networks as a trend for 2022, given it was mentioned in last year’s post. But while a year ago we foresaw the rapid acceleration towards zero trust network architectures, we now believe it to be a default approach. The COVID-19 pandemic has played a role here too, as much more flexible work has seen more devices previously used within the organization’s walls connected remotely over the public internet.
When zero trust networks means that the security profile for each device and application connecting to a network is independently evaluated each time it connects, it has significant implications for the video surveillance sector. Signed firmware, regular software updates, secure boot, encrypted data/video and secure identity will become hygiene factors in customer solutions, moving from ‘nice to have’ to ‘must have’.
Authenticate everything
While taking a zero-trust approach to cybersecurity is focused on authenticate the credentials of connected devices and applications, the ability to establish the authenticity of video surveillance itself is fundamental to trust in its value.
Tampering with video after it is captured, along with the increased sophistication in creating manipulated images, means that we may see the authenticity of video surveillance footage being more regularly questioned.
It is therefore going to be imperative that video surveillance can be undeniably established as genuine. Our approach is to add a digital signature into the video stream at the point of capture – a hash in each video frame – supplying proof that the video was produced inside a specific camera and that it hasn’t been tampered with since.
But this is an issue for the security industry as a whole. It is therefore imperative that the industry aligns behind initiatives to standardize on approaches to secure the authenticity of video footage captured by video surveillance cameras, ideally based on open-source software and initiatives.
AI becomes established and accepted (with appropriate controls)
It seems impossible to write a technology trends post without mention of artificial intelligence (AI). Many would also argue that AI is no longer a trend, and that the AI ‘genie’ is well and truly out of the bottle. Indeed, we’re all using and being exposed to valuable AI and deep learning-based services on a daily basis.
Our view remains that technology in itself shouldn’t be regulated, but that use cases of new technology should be. Legislation and regulation relating to the development and use of AI-based technologies and applications should be developed at a local, regional and international level. And it goes without saying that it should be adhered to by every organization employing AI.
While we’re still positive about the potential for AI and deep learning in video surveillance, we expect to see an even greater focus on initiatives to ensure that AI is being implemented ethically and without bias.
This is a positive thing and will be even more important as AI becomes embedded into every aspect of video surveillance. Greater integration of AI into the most fundamental levels of technology – the system-on-chip (SoC) – will see AI employed to enhance and optimize all aspects of video surveillance performance, from camera configuration to image quality to analytics.
COVID-19 as a catalyst
The long-term impact of the COVID-19 pandemic is being manifested in a number of ways. As mentioned in last year’s post, the pandemic has been a catalyst in low/no touch technologies, many of which are now embedded permanently, as is the use of intelligent video to ensure that social distancing and public health guidelines are being adhered to.
In relation to the technology sector, the pandemic also resulted in supply chain issues that have caused many organizations to consider how they create and source key components in their products.
The ‘connected’ nature of everything has meant that the global shortage in semiconductors has been a significant issue in many sectors, from consumer technology to automotive manufacturing. This in turn has led to more organizations – Tesla, Apple and Volkswagen among them – publicly stating a desire to design their own semiconductors, or system-on-a-chip (SoC) (though it should be stressed that designing an SoC and manufacturing it are very separate activities).
While this might represent a trend in some sectors, it is of course something that Axis has been doing for years with ARTPEC, and certainly designing SoCs that are optimized for specific applications is something that we anticipate more organizations doing in the security sector and beyond.
5G finding its place
Some might think we’re behind the curve in highlighting 5G as a ‘trend’ in the surveillance sector given that it’s been high on the agenda for a few years. But we see a fundamental difference between ‘hype’ and a ‘trend’. For us, a new technology only becomes a trend when we start to see valuable use cases appear in the security and surveillance sector. Though we still think it’s early days, this is starting to happen with 5G.
While much of the hype around 5G has been focused on improvements in network performance for consumer applications, one of the more interesting areas is how private 5G networks are emerging as a more compelling use case for the technology.
We do feel that private 5G networks show some genuine potential for video surveillance solutions across large or multiple customer sites and could bring particular benefits from a cybersecurity perspective. Certainly, if customers are creating private 5G networks then video surveillance will need to integrate seamlessly. Watch this space.
All trends viewed through the lens of sustainability
Sustainability can no longer be considered a trend. It needs to be embedded in everything we do: how we design and manufacture products, how we run our business, the performance of our suppliers – all aligned to reducing our environmental impact and operating in an ethical and trustworthy way. Wherever a technology trend looks like presenting an opportunity, it needs to also be examined through the lens of whether it can be developed and brought to market in a sustainable way.
From the power efficiency and materials used (and re-used) in a camera, to where and how it is manufactured and delivered, to the ethical implications of new technologies and business practices, interrogating trends against sustainability criteria is a critical as identifying them in the first place.
2022 will no doubt be another fascinating year; one not without challenges but which will also bring significant opportunities. As ever, we’re looking forward with optimism.