In recent years, we’ve been steadily adjusting to the introduction of IoT devices into our workplaces, vehicles and homes. By 2025, it’s predicted there’ll be 41.6 billion IoT devices in deployment globally. 

Organisations in particular have discovered how IoT technologies open up new ways to manage, monitor and measure assets and their performance. This has the potential to uncover valuable insights that would otherwise have remained hidden.

Organisations and society are now ready for the next evolutionary step – scaling up individual ecosystems into smart city deployments where all connected devices gather and share data to enable a new vista of possibilities.

As the number and type of connected devices rises, the management burden grows. It takes more time and expertise to configure and manage IoT solutions, as well as to analyse the results and decide what actions to take.

This will soon become unsustainable. Without automation, there’s a distinct danger that the huge management burden will end up outweighing any advantages larger IoT ecosystems might bring, causing development to stall.

It’s vital we recognise the importance of creating and building in automatic capabilities, known as ‘autonomics’ in IT management. These capabilities of self-configuration, self-healing, self-protection, self-optimisation and self-adaptation will protect the future potential of IoT ecosystems.

We’re investing heavily in making this future possible. Our involvement with pioneering research projects like our recent trial of a smart port ecosystem is just one example. 

This testbed showcases the seven key breakthrough benefits autonomics can deliver to IoT ecosystems:

Traditional IoT onboarding isn’t viable when millions of devices are involved. During our smart port trial, we tested an easy device onboarding system that installed thousands of IoT devices securely at once. It took care of: 

• locating and connecting devices on the network
• configuration and checking actions against the relevant provisioning policies
• running its own updates – all without human input.

Larger IoT ecosystems have an expanded attack service and increased vulnerability, and the consequences of a malicious attack on a smart city or smart road network could be catastrophic. 

We trialled a robust remote attestation system. This used automation to remotely authenticate IoT devices as they connected to the network, specifically requesting device characteristics that can’t be replicated.

Smart cities will need to use smart cameras with computer vision technology to detect objects, monitor security and improve health and safety. But there are privacy implications with this level of non-stop surveillance at scale. We successfully installed an automated video tool. This automatically anonymised faces and used GDPR-compliant analytics to identify potentially sensitive or private content.

Smart environments must be safe if they’re to take off - particularly in ecosystems with inherent dangers, like industrial sites or smart roads. 

During our trial, we experimented with smart cameras that continuously monitored environments using computer vision analytics. These detected health and safety issues like workers falling or wandering into dangerous areas, and were able to raise the alarm faster than a human could.

A large smart camera network will generate a constant stream of visual data that will need to be processed, securely stored and then viewed. There may well be latency challenges that come from relaying footage across cities, or even countries, so it'll be critical to cache and process at the edge to ensure operational efficiency.

We’ve been working on a highly scalable, high availability video storage platform that’s built on Apache Cassandra. It can automate video feed exchange to send footage securely between the cloud or edge, and is already integrating well with other platforms. 

When so much of a smart city or ecosystem depends on a continuous service, it'll be vital to detect and fix outages, viruses or misconfigurations quickly. The best defence will be a combination of automated network health monitoring, network micro-segmentation and policy-based incident response. 

These IoT autonomics continuously scan the ecosystem and isolate any devices responsible for unusual activity into their own micro-segment. Once there, they’re remotely scanned and remediated, before being securely brought back into the wider ecosystem. 

Larger ecosystems will have a variety of different groups of people operating inside them, and they'll be using them in different ways. To meet these diverse requirements during our smart port trial, we explored how an automated management system could generate tailored service level agreements for different ecosystem tenants. This included things like more accurate pricing based on space usage or device activity data. 

Our system measured contextual data against specified performance indicators to ensure service agreements were met. It also explored where operations, space or pricing could be further optimised. 

IoT autonomics have huge potential – enabling new environments that are smarter, safer, healthier and more efficient. If you’d like to find out more about our work as a leading innovator in the IoT space, download our whitepaper on automation in the Internet of Things. 

This blog post is based on the research, development and experience of our world-leading research and network strategy team at Adastral Park, who are continuously driving innovation through collaboration with technology partners, industry bodies and academics, in pursuit of tomorrow's technologies.