The broadcast industry has seen major transformations, but what we’re facing now is a deeper shift. It’s not just transport or tools. It’s a rethinking of how we build, scale, and operate live production infrastructure.
You’ve probably heard “infrastructure as code” a lot lately. For some, it’s the next step in automating and scaling live production. For others, it sounds like handing your OB van to a DevOps engineer with a YAML file.
I get the skepticism. But if you look at where the pain points in live production are today, you’ll see this isn’t just another tech trend. It’s a strategic necessity.
Why This Feels Risky (And Why It’s Necessary)
For decades, broadcast engineering has been a hardware-first, hands-on craft. SDI routers, frame syncs, multiviewers, and patch panels were built with precision and meant to run 24/7. Rebuilding a live environment on demand in the cloud? That’s a radical shift.
This works well for customers in sports, OTT, and fast-channel environments. They want modular workflows, not one-size-fits-all systems. But flexibility brings complexity.
It’s relatively easy to code infrastructure to spin up servers, storage, and connections. But configuring tech from 10 different vendors? That’s the real challenge. You need a unified, software-based control system.
The Real Bottleneck: control
The industry’s made progress solving video transport. ST 2110 helped us move beyond SDI into scalable IP production. But in focusing so hard on transport, we overlooked something else: control.
As we added more vendors, cloud services, and hybrid workflows, that gap became more obvious. It’s not just the complexity. It’s the manual overhead of stitching together multiple systems in real time. That doesn’t scale.
Customers don’t want one big system from one vendor. They want to curate their own stack. But large, complex production venues could involve over 200 control protocols across their equipment. No one would design a system that way, but here we are. Event-based, ephemeral infrastructure makes sense. But with a dozen vendors, each with different APIs and quirks, spinning it up isn’t trivial.
What is Infrastructure-as-Code?
It means managing infrastructure with software logic instead of manual steps. Think:
- Configuration as version-controlled text files
- Automated deployments of full production environments
- Predefined templates that launch workflows in minutes
Engineers don’t need to become Python programmers. But they’ll spend less time fixing protocol mismatches and more time producing content.
In software, Infrastructure-as-Code defines entire environments, networks, services, storage, application workflows, via configuration files. Tools like Terraform and Ansible aren’t just for DevOps anymore. Broadcast engineers need them to deliver scalable, on-demand infrastructure.
At AWS, this approach runs some of the most advanced live workflows in the industry. It’s not just about cost, it’s about agility.
Not Just Hype: A Culture Shift
Let’s be honest: not everyone’s ready for this. Broadcast engineers aren’t usually trained in cloud orchestration or declarative configs. That’s not a criticism. It reflects how the industry grew up. Our workflows were built to be robust and hands-on.
Going in as a traditional engineer will be tough. If we don’t invest in training — or use automation to bridge the gap — adoption will stall. But tools exist to help. That’s part of what ST 2138 (formerly known as Catena) is about.
ST 2138 is the latest standard being proposed by The Society of Motion Picture and Television Engineers (SMPTE). It provides a vendor-agnostic, software-based control layer across on-prem, private cloud, and public cloud, without rewriting APIs each time.
The Payoff: Real Agility
The goal isn’t to turn live production into a science experiment. It’s to reduce friction—especially when you need to launch fast, reconfigure often, and scale on demand. Software-defined infrastructure lets teams work faster and more consistently. It’s the only path forward if you need to do more with less. You can’t afford brittle, inflexible systems.
With a software-defined control layer you can:
- Launch events in hours, not days
- Automate failover, resource allocation, and monitoring
- Test environments in isolation
- Run infrastructure only when needed
Change is Hard, Standing Still is Harder
ST 2138 provides a unified control standard across vendors, platforms, and environments. It simplifies control and enables multi-vendor automation. This shift challenges legacy thinking, skills, and comfort zones. But it’s already happening. If you’re skeptical, that’s healthy. But the transition won’t wait.
To future-proof operations and protect quality and creativity, we must start seeing infrastructure as code: reusable, testable, scalable, and under your control. That’s not the end of engineering. It’s the evolution of it. Infrastructure as code won’t replace the art of live production. It just changes how we manage the canvas.
The draft of ST 2138 is now open for consultation. Click here to review the draft and join the conversation.
