CTO at Axis Communications
As Greek philosopher Heraclitus said, “the only one constant in life is change”. And this is certainly true for anyone working in areas related to or based upon technology (and few don’t these days). The pace of technological innovation is such that even the most fantastic of imagined futures seem like they could easily become reality.
As existing technologies reach maturity, unforeseen developments arrive ever more quickly, and innovations make the leap from consumer applications to business (and vice versa) it’s imperative that we constantly seek to find those that have the potential to add value to our own business and those of our customers.
As we look ahead to 2018, I’ve been working with my colleagues to identify some trends that we think will have an impact on our business and industry.
1. A move towards the edge
Two trends of recent years that have become familiar – cloud computing and the Internet of Things – have delivered undeniable benefits to businesses and consumers alike. But they also come with implications: namely the huge increase in the amount of data being transferred from connected devices to the data center for processing and storage, and the associated bandwidth needed. Edge computing alleviates this issue by performing data processing at the ‘edge’ of the network, near the source of the data. Doing so significantly reduces the bandwidth needed between sensors and devices and the data center. A further drive towards edge computing relates to potential concerns around data integrity and privacy: anonymizing and creating encrypted data within the device at the edge before it is transferred to the data center will be a likely response to these concerns.
As network cameras, audio and other sensors – the devices on the edge of the network –become ever more sophisticated and of higher quality, the need to balance both cloud computing and edge computing domains will be imperative to deliver refined, reliable and usable data (and for a an understanding of the differences, see here).
Despite the move towards edge computing, as outlined above, cloud computing will still play a significant role in IT infrastructures. But while cloud computing can give the impression of a single entity, there are, of course, multiple ‘clouds’ being used around the globe. As an increasing number of companies offer cloud-based services, the cloud eco system is increasingly becoming the preferred point of integration, rather than the traditional on-premise systems.
One benefit of integration between clouds is a significant potential reduction of in-house IT services required. Further, however, advanced composite services from multiple providers can be created and deployed through rich service APIs, including data analytics, content management, and storage, reducing time-to-market and rapidly increasing scale. Any organization delivering cloud-based services should be investigating the opportunity for integration with related services to add value to customers and partners.
3. Deep and machine learning
We have now reached a stage whereby the full benefits of deep learning architectures and machine learning can start to be realized: we have huge sets of data to analyze, the processing power available to do so within reasonable timeframes, sophisticated algorithms, and a weight of use cases to learn from. When some of the most impressive demonstrations of the application of deep learning have been related to image interpretation, speech recognition, and decision support, the potential for analytics in the safety and security sector are obvious.
At a relatively basic level, deep learning applications will improve video motion detection, facial recognition (and area where Axis partner Herta is doing some leading work), individual tracking and suppression of false alarms. It will aid system design, configuration, optimization and device management. Beyond this, as applications develop, there is significant opportunity for predictive analytics leading to incident prevention: from terrorist incidents to slip and fall accidents; from traffic issues to shoplifting; and even, as covered here, the tragedy of rail suicides.
It is still early days, however. Development is currently fast and unpredictable, and the demands on processing power are massive, but the potential for deep learning, which may ultimately lead to autonomous systems, is huge.
4. Personalization vs. privacy
One of the potential applications for deep learning could be in the delivery of highly personalized services. Imagine a retail environment where a customer’s face is recognized upon entering a store, and then offers are pushed to their mobile device based on previous purchases, preferences, or even their recent browsing history. But then, just because something can be done doesn’t necessarily mean that it should be, and this example immediately highlights increasing concerns around privacy, and how personal data is being used by businesses and other organisations.
Legislation is being created to address these concerns. In the European Union, the forthcoming General Data Protection Regulation (GDPR) – the deadline for compliance with which is May 2018 – will unify the protection of data for individuals within the EU, wherever that data is held or used.
But whether motivated by legislation or simply wanting to do the right thing by customers and citizens, balancing increased personalization with the protection of an individual’s data and privacy will be a tightrope all organizations will walk this coming year.
Once again, as it did last year, cybersecurity must appear on the list of trends for the next 12 months and beyond (if you’re keen to get under the bonnet of cybersecurity, you can find plenty of further reading here). The constant enhancement of cybersecurity will be a never-ending task, because well-resourced cybercriminals will never stop looking to exploit vulnerabilities in any new technology. And as the number of connected devices grows exponentially, so too do the potential flaws that, if left unaddressed, could provide the opportunity for networks to be breached, ransomware to be planted or, more simply, costly downtime to occur. 2018 will no doubt see more attacks and vulnerabilities exposed. The answer is proactivity and a systematic process for ensuring that patches are implemented as soon as they are available (our Hardening Guide is a useful resource).
6. Platforms to realize the full benefits of IoT
Speaking of the IoT (and I’ve looked at its implications for security previously) it has reached a point where to scale, collect and analyse data, and manage the network of connected devices effectively, it is crucial to use a scalable architecture. Such a so-called IoT platform allows equipment from different node vendors to coexist and easily exchange information to form smart systems using existing network infrastructure. There are numerous companies, both well-established providers of technology and new market entrants, enabling platforms to support IoT devices, and the next year will see further maturation. However, what will also be important for the future will be new international or de facto standards to enable interoperability between the different IoT platforms, and which will support true vendor-agnostic systems.
7. The blockchain: more than Bitcoin
For many, blockchain and Bitcoin have become synonymous. In reality, they are quite separate and while Bitcoin uses blockchain as its foundation, the potential for blockchain to verify almost anything that has a value is almost limitless. As an open, distributed ledger that can record transactions between two parties efficiently and in a verifiable and permanent way, the coming year will start to see blockchain being tested in multiple applications across numerous sectors.
In our industry, given that blockchain enables the authentication of any content, blockchain could be used to verify video content from multiple sources – such as public mobile phones and law enforcement body-worn cameras – for use within forensic investigations. Beyond video data, blockchain could also be used to verify the authenticity of devices connected to the camera network.
8. Breaking down smart city silos
The concept of a smart city (and safe cities as a result) isn’t a new one. For several years, the increasing number of sensors of different types being placed throughout urban environments are helping to solve specific use cases, from law enforcement to monitoring air quality. As the world’s population is increasingly housed in cities – 25% more people will live in a city in 2050 than do today – the use of sensors to help create more liveable, sustainable and safe environments will only increase.
However, a true smart city is a vision for urban development that integrates information, data, communications and Internet of Things (IoT) technology in a secure fashion to manage a city’s assets. These assets include government departments’ information systems, schools, libraries, transportation systems, hospitals, power plants, water supply networks, waste management, law enforcement and emergency services, and other community services.
Traditionally, most of these individual services have operated in silos. And this is holding back the realization of the vision for smart cities. A city can only truly be ‘smart’ when all of its data is open and usable across every service. Dealing with urban challenges such as safety and security for citizens, traffic congestion, aging infrastructure, and responses to events such as natural disasters and terrorist attacks, requires coordinated analysis of available data in order to deliver an appropriate and effective response.
9. Non-visual sensors bring new dimensions
Until recently, the primary – if not only – data available to surveillance operators was video which obviously only delivers a two-dimensional perspective. With the use of new, non-visual sensors, this view will become multi-dimensional, providing a richness of data that will enable for more rapid and accurate assessment of situations and, therefore, faster escalation, the activation of an appropriate response and the minimization of false alarms.
Radar technology, for instance, uses electromagnetic waves to detect movement. Radar is not sensitive to the things that normally trigger false alarms – like moving shadows or light beams, small animals, raindrops or insects, wind, and bad weather – and it can provide detail about any object’s exact position and direction of travel. And while thermal imaging is a non-visual technology already relatively well-established, advances in the accuracy of sound detection, whether a window smashing, or raised, aggressive voices – mean that audio will bring another useful input that may be missed in a purely video-based solution.
10. Virtual assistants and augmented reality leap into business
The last year has seen significant consumer adoption of virtual assistants. Amazon Alexa, Google Home, Apple Siri, and Microsoft’s Cortana have all gained momentum as technology to help people manage their daily lives, and forthcoming technologies such as Facebook M will only build on this. It is inevitable that these same technologies will start to find their way into the business environment, as consumers expect the same levels of technological help at work as they now get at home. Particularly for the providers of any sophisticated or complex technology-based products and services, virtual support in specification, installation, configuration, and management will become more than simply expected; it will become an imperative.
Similarly, augmented reality (AR) is a technology that has largely existed in specific niches to date, such as the military and aviation, but again shows enormous potential in the business arena, particularly now it is available as a technology in mobile devices and in an increasing number of wearables.
One of the obvious opportunities for AR in the business arena is in the installation and maintenance of technological solutions, where visual instructions can be overlaid on the real-world view of technicians to aid them in the work. In specific relation to our own business, with the growing use of non-visual sensors and analytics to add accuracy and further perspectives to visual information, consumers of video surveillance will be using AR to bring these sources of data together in a single view, enabling a more rapid and appropriate response.
So, there you have it: 10 trends we at Axis think will shape 2018. But these are only our views. Do you agree or see another technological development that will have a significant impact?