DTW24-Ignite: Telcos and hyperscalers adopt ODA Canvas for plug-and-play interoperability

This week at DTW24-Ignite TM Forum announced that the Open Digital Architecture (ODA) Canvas, an execution environment for telecoms IT and network components, will be generally available for wide-scale use in January 2025. Communications service providers (CSPs) and cloud platform providers are already using it to integrate and deploy IT components in a single, secure cloud-native environment, and longer term the Canvas will make it easier for them to move workloads from one cloud to another.

Development of the ODA Canvas began in 2019 as part of a TM Forum Catalyst project called the Business Operating System, which aimed to produce an interoperable reference implementation of a telco’s core commerce management system including a product catalog and order management service. The point was to drastically simplify integration of IT systems by developing a plug-and-play environment to enable zero-touch operations.

Now, that vision is becoming reality. “It’s not theoretical; it’s not paperwork. We are creating a reference implementation that is based on Kubernetes,” says Sigrid Braun, an Enterprise IT Architect at Deutsche Telekom who is active in TM Forum’s Innovation Hub where the reference Canvas is being created.

“The new collaboration is done by an Innovation Hub team with DevOps engineers, platform engineers and architects who are creating real software for the reference implementation together,” Braun explains. “With this approach we are much faster to use the new standard because there is already a reference implementation available that we adapt to our Deutsche Telekom needs.”

Customizing canvases

Indeed, DT is developing an ODA-compliant canvas called Magenta, and other CSPs such as Bell Canada, Jio, Orange and Vodafone are also creating their own canvases where ODA Components, the software building blocks of a CSP’s cloud-native IT and network systems, can be deployed.

“The Magenta Canvas provides a standard way to deploy and operate TM Forum-compliant software components securely on different cloud infrastructure integrated with DT applications,” says Braun, adding that the company is looking to enhance the experience of developers with its canvas. “To increase the productivity of our DevOps teams, the Magenta Canvas will be powered by our Developer Experience Portal, which simplifies the access and use of the canvas,” she explains.

Vodafone’s canvas is its Telco-as-a-Service, or TaaS, platform. “TaaS already exists as a way for different markets and functions within Vodafone to effectively be a tenant of a Kubernetes platform. It is used by our Site Reliability Engineering teams to host mainly our digital engagement systems, which run on Kubernetes,” says Dr. Lester Thomas, Head of New Technologies and Innovation at Vodafone Group, and leader of ODA Canvas development in the Innovation Hub. “But this canvas adds a layer on top so we can run core IT systems… Today, those systems typically run in an older environment. And even if they’re running in the cloud, they’re running in separate, siloed cloud instances, not as modular components in a single platform. Creating a single platform is what the ODA Canvas does.”

An operating environment ‘at the click of a button’

Among hyperscalers, Google and Microsoft are both contributing to the development of the reference Canvas in the Innovation Hub, and each has also built its own canvas. Google is a founding member of the hub, while Microsoft joined the group in April. Amazon Web Services (AWS) has developed an ODA-compliant canvas as well but is not yet a member of the Innovation Hub.

Soon, CSPs that are not developing their own canvases will be able to go into a hyperscaler’s marketplace and purchase one that they can use to deploy components. Hyperscalers and ISVs are working on possible commercial models, with all three cloud providers, plus Oracle and Red Hat, demonstrating their canvases with CSPs and other partners at DTW this week.

Bell Canada and AWS, for example, are showing how deployment of ServiceNow components along with ODA-compliant components on the AWS canvas can address a big challenge that most CSPs are facing: “all the sidecars that we have to implement from one cloud to another, like the login for security and the audit logs for telemetry,” says Neel Mehta, Director IT Delivery at Bell Canada. “And each developer chooses his own different pattern of how they want to log the information, so the metadata may differ as well,” he adds.

Sidecar patterns are so named because they act like a motorcycle sidecar, attaching to a parent application to provide supporting features like identity management, API management and observability for monitoring. “So, what the ODA Canvas gives us is the ability to have this logging and security standardized in one canvas,” Mehta explains. “We don't have to worry about them… We just focus on the business processing and business workflows.”

The reference ODA Canvas provides open-source solutions for these sidecar component services, but cloud providers can offer their own methods as part of their compliant canvases.

Indeed, Microsoft’s vision for its canvas is “to automate the entire ODA Canvas to such an extent that at the click of a button, you will have an environment ready in which you can deploy your components,” says Jayant Mishra, Senior Industry Digital Strategist at Microsoft, who is a participant in the hub. “There is a lot of work to be done to get there. But having an entire GitHub repository – wherein we just go and do a pull request and then we will have a complete Kubernetes-based environment ready with all the security, monitoring, logging and API management in place in which ISV [independent software vendor] partners can deploy their components and integrate and test their components – that is the ultimate North Star.”

Adding AI to canvases

All three hyperscalers plan to deliver a full suite of managed services with their canvas platforms. Google, for example, is offering its API management service, Apigee, along with Service Mesh (ASM), Google Managed Prometheus (GMP) and Identity & Access Management (IAM) to help CSPs take advantage of the cloud provider’s data analytics and Vertex AI models.

“This provides a strong platform for telcos to host AI-first ODA components for their core business applications,” says Priya Saxena, Strategic Cloud Engineer at Google, and a participant in TM Forum’s Innovation Hub. As a founding member of the Innovation Hub, Google has been working with CSPs on how to integrate AI capabilities into the ODA Canvas.

“We want to look at what telcos can do with all these AI capabilities as traditional AI algorithms and generative AI are coming together,” says “How can we take cloud infrastructure to that next level, where telcos have readily available models to say, 'You know what, my billing system really works well with this fraud detection model,’ or ‘My product ordering system will work very well with this kind of risk analysis model’.”

Saxena adds: “We want telcos to take that extra step forward to consider how to modernize their BSS/OSS with this native-AI kind of approach. By integrating AI into the ODA Canvas, it’s not just a plug-and-play model for your infrastructure. It's also going to be a standardized model for your future-generation, AI-native OSS/BSS.”

Something for everyone

Ultimately, the ODA Canvas makes it much easier for CSPs to switch out components and move component workloads from one cloud to another, which ISVs and hyperscalers could view as positive or negative. While it means that telcos are not “locked in” to buying their products and services, it also means that a CSP that is not currently a customer could become one more easily.

For providers of canvases and components, the ability to differentiate their capabilities is key. “The standard makes sure that the platform can be deployed seamlessly and can interoperate seamlessly. But what that platform is and how is it developed – how efficient it is and how comprehensive it is – solely remains the creativity or the innovation of the platform owners,” says Dr. Sudhir Mittal, EVP & Chief Architect, Jio Platforms.

Jio, which hosts the Innovation Hub in Mumbai, wears two hats: It is both a large mobile operator serving 500 million subscribers in India and a cloud-native platform company that sells OSS/BSS components.

“What is inside a component is defined, but how that functionality is to be provided with what kind of additional features, with what kind of state-of-the-art AI capabilities or GenAI capabilities, that is really the richness of the component or platform,” Mittal explains. “Compliance is the basic, minimum thing, but what we give apart from compliance is a feature which provides a lot of flexibility and creativity to the user of that component. That’s important.”