How do you explain the edge in simple, easy-to-understand terms?
Ian Wadgin, VP technology and innovation at Appear: The edge is the venue, the stadium or the studio, and doing some media processing there can be so much more efficient and flexible than backhauling everything to a central data centre or cloud for processing. Think of it as taking the heavy lifting to the event, not the event to the heavy lifting. In practice, that means processing, compression, encryption, protection, and format conversion happen in the first mile, so what travels across the network is already secure and broadcast-ready. It is not a replacement for cloud; it is about putting each workload where it performs best. In some instances where connectivity bandwidth is constrained it may also be beneficial to have operational functions such as vision mixing or sound processing at the edge and controlled centrally.
It is possible to build a mini data centre at the edge of cloud architecture and then just send relevant outputs such as programme and multi-viewer feeds over the network, removing the need to send every source back to the central processing platform. As more functions become based on software rather than hardware, we can think about this as replacing a traditional OB truck with a data centre in a couple of racks or on wheels.
Francesco Scartozzi, Matrox Video’s VP of sales and business development: Think of the edge as the point where the digital production environment meets the physical world. Cameras, microphones, displays, and operators all exist at the edge. Those signals still need to be captured, synchronised, transported, monitored, and delivered with predictable performance before software can do anything with them.
While more production functions are moving onto standard IT infrastructure, the edge remains essential because it’s where real-time media enters and exits the workflow. The future isn’t about eliminating edge devices, but allowing edge infrastructure and software-defined production to work together seamlessly.
What are its advantages in live streaming?
IW: The most obvious advantage is latency. Processing at source removes round trips to distant data centres, which matters for live sport, betting-adjacent content, and interactive formats. The economics follow close behind: compressing at the venue with codecs such as JPEG XS or HEVC before transport cuts contribution costs dramatically and makes premium remote production viable over the internet rather than expensive leased circuits or satellite links. Then there is resilience. Deterministic, purpose-built processing at the edge, combined with protection such as ST 2022-7 and SRT, keeps output stable even when networks misbehave – and in live, there is no margin for error. It also protects quality by protecting the signal at its high point, before it ever meets the public internet.
FS: Live production is unforgiving. Every frame matters, and delays or synchronisation issues quickly become visible on air. Edge devices perform the time-critical functions closest to the source, allowing media to be captured, synchronised, encoded, decoded, and transported before it reaches software-defined processing environments.
They’re also what make modern remote and distributed production economically viable. By bringing media onto IP networks at the edge, broadcasters can transport high-quality feeds over long distances without relying on dedicated broadcast infrastructure between locations. That reduces the cost and complexity of connecting venues, centralising production, and scaling coverage to more events.
Which use cases do you feel present opportunities in terms of the most untapped potential?
IW: The biggest untapped opportunity is scaling remote production down, not just up. The same edge economics that serve tier-one events make tier-two and tier-three sport production commercially viable, a far larger market than marquee events. AI at the edge is an exciting development, with tools soon to be available to automate functions such as the mixing of audio and video, metadata tagging and highlights generated at the venue rather than shipping raw feeds to the cloud for processing.
There is also real potential in the venue-as-a-platform model, where stadiums with permanent edge infrastructure serve multiple rights holders simultaneously from a single contribution layer, and in distribution, edge processing at affiliate and regional level that enables faster onboarding and localisation without central re-engineering making more feeds available for a richer content experience. With edge infrastructure it is not necessary to have all the kit or crews travel to the venue. By splitting control and processing, it is possible to redeploy staff and production facilities across multiple events in the same day or weekend.
FS: One of the biggest opportunities is enabling organisations that previously couldn’t justify broadcast-grade infrastructure. Regional sports, collegiate athletics, corporate media teams and specialised content producers all want to deliver higher-quality productions while operating with leaner teams and tighter budgets. Software-defined production combined with edge infrastructure lowers the barrier to entry, allowing these organisations to build workflows that would previously have required far greater investment.
At the same time, larger broadcasters are using similar technologies to launch new FAST channels, support remote production, and extract more value from existing infrastructure. In both cases, the goal is the same: make production resources more flexible and better utilised.
How can security risks be overcome?
IW: It starts with treating every edge node as part of the attack surface from day one; security must be designed in, not bolted on. Encryption in transit should be standard, over SRT and similar, with key management that works operationally under live conditions. At the device level, hardware-based security and signed firmware matter, alongside network segmentation so that a compromised link cannot reach the wider production network. Just as important is disciplined lifecycle management: patching, access control, and audit across a distributed fleet, applying structured, auditable security processes to broadcast infrastructure, rather than ad hoc security fixes. This ensures that security is treated as a continuous process (including risk assessment, patching cadence, access control, audit logging) rather than just set and forget practice. Availability is a security property too, redundant, diverse paths such as ST 2022-7 protect the output as surely as encryption protects the content.
FS: As broadcast adopts more IT technologies, security naturally becomes a bigger consideration. The good news is that broadcasters don’t have to start from scratch. They can draw on well-established IT practices while designing workflows that meet the reliability demands of live production.
And what about computing limitations?
IW: The edge is a constrained environment by default. Power, space, and cooling at a venue are not as efficient or as powerful as at a data centre, so efficiency per rack unit and per watt is the real currency. This is where purpose-built hardware earns its place: dedicated processing delivers deterministic, low-latency performance at a density and power envelope that generic CPU or GPU compute cannot match. Density also matters commercially; the more channels and services per chassis, the fewer boxes to ship, power, and manage at each location. Ultimately the optimal answer is hybrid: with deterministic and powerful media processing at the edge, elastic and bursty workloads handled in the cloud.
FS: General-purpose compute has become remarkably capable, but live media still has unique requirements. Processing uncompressed video, maintaining precise synchronisation, delivering deterministic latency, and ensuring redundancy at broadcast scale remain highly specialised challenges.
The question isn’t whether IT infrastructure is powerful enough. It’s how to make standard IT infrastructure behave like broadcast infrastructure. That’s where software-defined frameworks, media-aware orchestration, and purpose-built edge devices complement one another. Rather than replacing specialised hardware entirely, we’re seeing production functions migrate to software while media acquisition, transport, timing, and connectivity continue to live at the edge.
There is an ongoing narrative around broadcast becoming an IT industry. To what extent is the tail now wagging the dog in terms of technology driving what happens in production?
IW: The tools have changed; the job has not. IP, COTS, and cloud are means to an end – the programme, the audience, and the live moment remain the point. It is also worth remembering that broadcast has not simply adopted IT; it has bent IT to its requirements. ST 2110, PTP timing, and hitless protection exist precisely because generic IT could not meet broadcast’s determinism and availability demands. That said, the risk is real that technology roadmaps start dictating production decisions rather than serving them. The healthiest framing is that production defines the outcome and engineering defines the path. When that ordering holds, the broadcast dog is still firmly wagging the IT tail.
FS: I think it’s important to distinguish between the technology and the reason the technology is being adopted. Broadcast has evolved continuously over the years, and each major transition has been driven by the industry’s need to respond to new business realities. Today, broadcasters are under pressure to produce more content, reach more platforms, and create new revenue opportunities while operating more efficiently. Traditional broadcast infrastructure wasn’t designed for that level of flexibility and scale, so the industry has increasingly turned to IT as part of the solution.
That doesn’t mean broadcast is becoming an IT industry. Broadcasters still care about the same things they always have: quality, synchronisation, resilience, and operational reliability. IT simply provides a more flexible foundation for delivering those outcomes.
- This article originally appeared in the September issue of TVBEurope, available to download here.
