AIoT needs the network—and the network needs AIoT

Hello readers,

Welcome to the IoT For All newsletter! This week we’re talking about how network technological and AIoT are intertwined, the benefits of eSIM technology for OEMs, retrofitting existing machinery for IoT and more!

Also, an important update to the IoT For All newsletter family: We’ve brought our AI and IoT newsletters together under one roof. That means more insights, fewer signups, and you’re automatically included.

The Backbone and the Brain: Why AIoT and Connectivity Go Hand in Hand

The networks of yesterday can’t quite cut it anymore. Thousands of IoT devices, mountains of AI data—the amount of information that needs to cycle through a typical enterprise network is already pushing the limits of bandwidth and speed. And that is just current demand. The infrastructure of the broader internet needs an upgrade, and connected enterprise operation centers will need to be on the bleeding edge.

What will this new backbone need to look like? It’s already coming into focus, and there isn’t just any one answer.

You can catch a glimpse of the future with the booming market for private 5G. An estimated $2 billion market in 2023, some projections predict growth of more than 50% to reach $36 billion by 2030. With the ability to support as many as 1 million devices per square kilometer, sub-millisecond response times, and resilience in noisy environments, 5G is a critical enabler for mission-critical IIoT and automation in scenarios where downtime can cost millions. It’s no wonder we’re seeing it installed at smart cargo ports, mining sites, and beyond.

Of course 5G is not a networking panacea. Accessing the full power of a private 5G deployment is complex technical work; not every IT department is up to the task, and the people who can are in high demand. The integration of a new cellular network with existing IT and operational technology systems adds layers of complexity, and plenty of opportunities to trip up.

Which is why the emergence of Wi-Fi 7 is another important ingredient in supporting high-density, AI-drenched IoT deployments—particularly the ones that have to live indoors. Its flagship features, like “wider” 320 MHz channels, advanced transmission mechanisms like 4096-QAM modulation, and ability to maintain multiple connections with Multi-Link Operation (MLO), provide up to four times the capacity and data rates compared to Wi-Fi 6.

Raw speed is thrilling, but particularly important are some of the unsexiest bits. Wi-Fi 7 focuses on “deterministic performance," the ability of the network to deliver data with predictable speed and low latency, even in high-density or congested environments. Likewise, Wi-Fi 7’s Target Wake Time (TWT) and Enhanced Target Wake Time (eTWT) which allows connected devices to determine their own wake up times will help ensure that battery-powered devices have the power to boot up and communicate when they need to.

While it’s tempting to think of these technologies as essential for enabling an AIoT future, where thousands if not millions of devices are transmitting huge data payloads nonstop, the relationship also goes in the other direction; the type of powerful hybrid network AIoT requires will not be able to exist without its oversight. The sheer volume and velocity of data in a hybrid Wi-Fi 7 and private 5G environment make manual network management impractical.

Seamless handover and traffic steering are crucial for a truly converged network. AI-driven network selection mechanisms ensure optimal connectivity based on real-time conditions, allowing seamless roaming between indoor Wi-Fi and outdoor 5G spaces. Technologies like Access Traffic Steering, Switching and Splitting (ATSSS), enable a device connected to both 5G and Wi-Fi to steer traffic across both connections, effectively "bonding" them into a single session.

As ever, the path to the network of the 2030s is not crystal clear from here in 2025. The wireless landscape is dynamic. Monitoring advancements in Wi-Fi 7 and 5G standards, as well as the looming arrival of 6G, will be crucial to finding the mix of technologies that can meet the moment.

One thing is for sure: AI may need the network, but the network needs AI too.

📖 Top Articles

SGP.32 and eSIM: Powering the Next Wave of IoT Innovation for OEMs

For developers, module makers, and OEMs operating in an endlessly dynamic IoT landscape, arriving at a scalable, secure, flexible, and cost-effective connectivity solution is absolutely crucial. Traditional SIM cards, which risk locking devices to one MNO, or running up roaming charges, only add more complications to already complex global deployments. Add in the manufacturing complexity of multiple SKUs, physical vulnerabilities, and form factor limitations, and it’s easy to see why anyone managing large IoT fleets would want a better solution. Fortunately, eSIM technology spearheaded by the robust GSMA SGP.32 specification, smooths out the speed bumps, empowering OEMs to redefine their IoT strategy with less logistical constraints.

🔥 Rapid Fire

• Elliptic Labs and Ceva collaborate to enable smarter edge devices

• Nokia taps AMD processors to power next-gen 5G networks

• RAKwireless launches $6.50 LoRa, Wi-Fi, and Bluetooth RAK11160 IoT module

• Heltec Automation partners with Morse Micro to advance IoT with Wi‑Fi HaLow

• Introducing the first-ever utility meter to satellite communication

🎙️ The IoT For All Podcast

In this episode of the IoT For All Podcast, Anthony Protopsaltis, Principal at Velocity IoT, joins Ryan Chacon to discuss how AI is reshaping IoT. The conversation covers the importance of resilient infrastructure, planning an IoT deployment, strategies for seamless communication and long-term reliability, the complexities of IoT connectivity, SGP.32, multi-IMSI, NTN, and advice for companies aiming to future-proof their IoT connectivity.

YouTube • Spotify • Apple • YT Music • Amazon