Across industries, more organizations are turning to the cloud for AI-driven video analytics. The promise is powerful: more accurate insights, the scale to run advanced AI models, and the ability to combine and analyze data from many sources for richer analysis.
But there’s a catch. Cloud analytics depend on the quality of the video they receive – and transmitting that video with all the detail analytics rely on is easier said than done.
To better understand this challenge and how new technology can help, we sat down with Mats Thulin, Director of AI and Analytics Solutions at Axis, and Stefan Lundberg, Senior Expert Engineer at Axis.
Mats brings the analytics perspective, focusing on how detail and quality drive better results in the cloud. Stefan has been a driving force behind technological advancements at Axis since the inception of the first network camera.
Stefan played a pivotal role in developing the ARTPEC chip family, Zipstream, and other core technologies that continue to shape video surveillance and analytics today. Together, they explain why AV1 codec matters for cloud-based analytics.
Managing bandwidth and data volume is one of the biggest hurdles in cloud-based video analytics. That’s where AV1 makes a significant difference. As a next-generation codec, it can dramatically reduce bitrate without compromising on the visual detail analytics depend on. That makes large-scale cloud analysis possible without overwhelming infrastructure.
The cost of lost detail
“Analytics thrive on rich visual information,” Mats explains. “The clearer the input, the better the algorithms perform. However, UHD, high frame rates, and HDR create massive data volumes. If you compress them with legacy codecs, either bandwidth costs explode, or image quality degrades – and both limit the usefulness of analytics.
This becomes especially critical in advanced use cases like running large-scale forensic searches, training custom AI models on rare events, or monitoring quality in complex industrial settings.”
H.264, still the industry’s most common codec, requires significantly higher bitrates to preserve clarity. Across hundreds or thousands of cameras, those demands quickly become bottlenecks.
From edge to cloud: why compression matters
Axis has long championed edge analytics, where raw or lightly processed video is analyzed directly on the device. This approach ensures that detail is retained, and latency is avoided.
“But when you want to aggregate results, enrich them, or run heavier AI models in the cloud, video quality on the way up becomes critical,” Mats notes. “You need an efficient compression method that still preserves detail. Otherwise much of the information is lost before it ever reaches the cloud.”
Why AV1 changes the equation
“As an engineer, I focus on both efficiency and quality”, says Stefan. “With AV1, the gains are clear. You typically see 40 percent lower bitrates compared with H.264. In practice, that means you can transmit the same quality at much lower cost or preserve much more detail within the same bandwidth.
For H.265 the numbers depend more on the scene and on which H.265 encoder you’re comparing with. A reduction of 25% compared with our own previous generation is typical even though they use roughly the same bitrate on ARTPEC-9”.
The amount of detail preserved makes a critical difference for analysis. “With H.264, once you push the bitrate down, the picture quickly starts to break apart,” Stefan explains. “You lose sharpness, small objects blur, and text or license plates become unreadable. AV1 is much better at keeping that information intact, even when bandwidth is limited.”
From the analytics side, Mats emphasizes the impact: “AV1 makes a real difference in scenarios where advanced analytics need every bit of detail. That could mean applying tailored AI models in logistics or factory environments or enabling granular forensic searches across large video archives.
Edge analytics will remain essential for tasks like real-time detection on the device, counting, and standard search. But when workloads shift to the cloud for more compute-intensive or large-scale analysis, AV1 ensures that the necessary detail is preserved without overwhelming infrastructure.”
Mats also highlights that AV1 codec is ready for practical use: “This isn’t just theory. AV1 is already running at scale, and its wide support and open-source model give organizations confidence that they can use it for analytics today – and in the future.”
Preserving the value of great imaging for analytics
Axis cameras are designed to capture the best possible clarity, even in challenging conditions. AV1 helps ensure that information doesn’t get lost on the way to the cloud. As Stefan puts it: “It’s about preserving image usability – making sure that the fine detail our cameras capture remains intact when it’s used for analysis later.”
Together with Axis innovations such as advanced imaging, Zipstream, and open integration platforms, AV1 strengthens the foundation for cloud-based solutions. The goal is simple: give analytics the best possible input wherever they run.
Looking ahead
Stefan highlights AV1 codec’s potential for the future: “Even at very high resolutions like 8K, AV1 can reduce bandwidth needs without losing picture quality. 8K may still be rare today, but it’s on the horizon – and with AV1, running those streams in the cloud becomes realistic.”
Mats concludes: “For analytics, the benefit is straightforward – more of the visual detail captured by the cameras makes it to the cloud without overloading the infrastructure. That’s what enables better insights at scale.”
In summary, cloud analytics depend on the quality of the input. With AV1, organizations can preserve the richness of video captured at the edge, ensuring that every stream contributes to better insights. By combining AV1 with Axis imaging expertise and open integration approach, organizations can scale AI-powered video analysis more efficiently in the cloud.
Learn more about how AV1 enables efficient video transmission for cloud analytics at axis.com/solutions/av1-codec.
For those looking to reduce bandwidth use through edge processing, visit axis.com/analytics for an overview of Axis edge analytics solutions.
