Autonomous, intent-based orchestration will allow the IoMT to scale

Commercial context

As the IoT evolves, it connects ever more mobile devices such as vehicles, robots and wearables. This 'Internet of moving things' (IoMT) is creating a rapidly expanding ecosystem with dynamic and unpredictable workloads. Services must be discovered and provisioned on the fly, and connectivity handed off continuously as devices move between network edge nodes. This fluidity stretches IoT management platforms’ capabilities. Ensuring service continuity across cells and meeting strict latency and performance requirements become much harder. Furthermore, the expansion of the IoT could drive up energy usage and carbon emissions if managed inefficiently. Traditional rules-based human driven operations can no longer address these challenges.

The solution

This is where the ‘Autonomous and sustainable IoT ecosystems at scale - Phase II’ Catalyst comes in. The first phase of this Catalyst developed a solution to follow a single 'moving thing' (a drone) hopping between different edge nodes to support IoT service and network orchestration. It also applied simple techniques for energy savings. The second phase focuses on supporting thousands of moving things - helping usher in a strategic vision for the future of IoMT applications.

The Catalyst is now developing a management platform that dynamically orchestrates networks and services for moving devices in real time. The team is leveraging event-streaming, and advanced machine learning, alongside intent-based networking, to automate the “follow-the-thing” process across distributed cloud and edge environments. In practice, this means that as a device moves or its service needs to change, the platform intelligently orchestrates the necessary network slices, edge resources, and IoT services on the fly, with minimal human intervention.

When a mobile IoT device approaches a new area, the platform will use APIs to discover the available IoT services at the nearest edge node. It then negotiates for the optimal network resources and will route the device appropriately. For example, if a drone in flight switches on a HD camera, the system anticipates the need for more bandwidth. It immediately orchestrates a suitable 5G network slice to guarantee high-quality video streaming. At the same time, the platform can clone and transfer other services, such as video analytics or an anonymization app, to the edge node closest to the drone, so data can be processed locally with low latency.

Behind the scenes, an AI-based orchestration engine interprets the device’s or application’s requirements as ‘intent’, which is then translated into concrete network actions. This AI orchestrator can juggle many devices, continuously optimizing routes and resources for each. Crucially, the solution uses AI and event streaming for on-the-fly anomaly detection and optimization, as well as orchestration. When the system detects an unusual pattern or fault (say a device losing connectivity abnormally or a spike in power draw), it automatically flag it or self-correct. For example, it could reroute the connection, adjust power settings or alert operators with recommended actions. This proactive AIOps (AI for operations) approach ensures the IoT ecosystem stays resilient at scale, identifying issues in real time before they cascade into service outages.

The Catalyst is also harnessing generative AI (genAI) to enhance decision-making and transparency. GenAI will derive actionable insights from the enormous amounts of data the platform aggregates. For example, a dashboard might suggest ways to re-route or delay non-critical tasks to off-peak times to save energy, or highlight underutilized resources that can be shut down. These AI-driven insights can enable autonomous decision-making to ensure the IoMT ecosystem runs in the greenest way possible.

Applications

For telecoms operators and enterprises, the solution unlocks innovative services and monetization opportunities and operational efficiencies. It can also create new revenue streams – every time an autonomous vehicle or drone requests a premium connectivity boost or edge service, that’s a chargeable event generating new income for an operator. Closed-loop feedback and automation also drives down operational costs for operators.

At the same time, businesses deploying IoT at scale benefit from assured performance (no dropped connections as assets move) for their mission-critical applications. The solution can also optimize costs by using network resources more efficiently.

Nektarios Georgalas, Innovation Principal Data and AI at BT and Catalyst project lead, explains. “Our Phase II Catalyst pushes the boundaries of IoT management. We’re showing that a swarm of moving devices can be managed autonomously at scale without sacrificing service quality or sustainability. The result is a glimpse of a future IoMT that runs itself. Where intelligent algorithms seamlessly handle connectivity and optimize energy usage.”

Wider value

This Catalyst intends to provide a blueprint to handle the next wave of connected devices. The goal is to create a sustainable IoMT where autonomous systems minimize the carbon footprint of connectivity. An intent-based, autonomous networking approach can extend to other verticals. For example, the automotive, transportation, shipping and agriculture sectors could deploy similar architectures to guarantee service continuity.

The industry will also benefit from the use of standardized APIs and frameworks in this solution. This means easier integration and faster time-to-market for new IoT offerings. By using open TM Forum standards, the Catalyst’s approach is replicable so others can build upon it, accelerating innovation industry-wide. More broadly, embracing autonomous IoT management at scale has clear societal benefits. For example, it can support new public services. These include drone-assisted emergency response or intelligent transportation systems that require always-on connections.

“Ultimately, this is about enabling operators to monetize 5G and edge services on-the-fly (think automatic network slices for drones or vehicles) while dramatically reducing carbon footprint through smart automation,” concludes Nektarios Georgalas of BT. “It’s incredibly exciting – we’re not just solving a tech problem, we’re demonstrating a whole new model for sustainable IoT business.”