How Smart Factories Are About to Get Smarter

In partnership with

Hello readers!

This week we’re talking about how the next frontier of Digital Twins enables Industry 5.0, the future of utility networks, smart farming, and 5G RedCap.

Digital Twins Grow Up — And Move to the Edge

Digital twins have been one of the most exciting ideas in smart manufacturing for nearly a decade, promising real-time visibility, better predictive models, and continuous optimization. But in practice, most digital twins depend heavily on the cloud, and that dependence creates the same bottlenecks every IoT team knows too well: latency, bandwidth consumption, and a constant tug-of-war between responsiveness and cost.

A new study from researchers in Taiwan revisits the digital-twin concept with a more modern blueprint, one designed to reflect how today’s factories actually operate. Their argument is simple but powerful: digital twins work best when intelligence lives where the action is: at the edge.

The proposed architecture runs both AI inference and physics-based simulations locally on factory-floor edge devices. Instead of shipping high-frequency machine data to the cloud every few seconds, the twin and the equipment stay synchronized on-site using standardized co-simulation techniques. For IoT teams this turns the edge into a fast, self-contained decision layer, capable of testing scenarios, optimizing workflows, and spotting anomalies in real time, without waiting for a round trip to the cloud. The factory essentially becomes a distributed compute cluster, simulating itself continuously.

The study also introduces a practical way to share model improvements across locations without exposing sensitive operational data. Using a federated-learning approach, individual factories train models on their own equipment data but share only anonymized updates. For industries that consider their process data proprietary, this design removes one of the biggest barriers to scaling AI beyond pilot projects.

What’s most striking, though, is the performance. In testing, the edge-oriented architecture cut latency by 35%, reduced cloud communication by nearly a third, and improved fault detection accuracy by 20%. In assembly-line benchmarks, it delivered over 13% higher throughput than cloud-only or edge-only designs. These are real, operational gains created by shifting simulation and analytics closer to the machines themselves.

The applications are easy to picture. Predictive-maintenance systems that react immediately instead of waiting for cloud results. Assembly lines that reconfigure on the fly when one station slows down. Multi-site operations that benefit from shared model updates without sharing any raw data. And factories where local digital twins run continuously, learning from the environment and adjusting parameters long before a human sees an alarm. It’s a move from automated systems to self-optimizing ones: the kind of shift that defines the transition from Industry 4.0 to Industry 5.0.

As always, edge-heavy architectures come with trade-offs: more distributed infrastructure, more coordination across sites, and new expectations for device-level processing. But the core idea lands cleanly: digital twins don’t need to live in the cloud to be effective. In many cases, they work better when they’re close enough to hear the machines they’re modeling. And if the edge becomes as capable as this research suggests, the next wave of smart manufacturing may be defined not by how much data factories collect—but by how quickly they can act on it.

📖 Top Articles

The Long View: What Enlit Revealed about the Future of Utility Networks

Walking the floor at Enlit in Bilbao this November, one couldn’t help but consider how the decisions made today about network infrastructure will shape how communities access energy and water for the foreseeable future. The LoRa Alliance had a significant presence at the event, with six members in its pavilion and more than 25 others throughout the venue. Companies like Janz, Kerlink, MultiTech, Semtech, TEKTELIC, and Zenner demonstrated utility solutions while countless conversations revealed both the opportunities and complexities utilities face as they modernize networks.

Read more

• How Software-Defined Connectivity Supercharges IoT ROI

• Smart Devices, Tougher Rules: The Future of Medical Device Compliance

• Smart Cities Are Getting Smarter—Quietly Racing Against Urban and Climate Challenges

• Demystifying the New eSIM for IoT: What Every Enabler Should Know

• Building Smarter Cities with LoRaWAN Connectivity

Free email without sacrificing your privacy

Gmail is free, but you pay with your data. Proton Mail is different.

We don’t scan your messages. We don’t sell your behavior. We don’t follow you across the internet.

Proton Mail gives you full-featured, private email without surveillance or creepy profiling. It’s email that respects your time, your attention, and your boundaries.

Email doesn’t have to cost your privacy.

Ditch the Gmail data grab

🔥 Rapid Fire

• Lacuna Space accelerating Brazilian AgTech alongside Embrapa and Wyld Networks

• Com4 launches next generation IoT connectivity management platform in collaboration with MAVOCO

• Sentient AIoT and fixed wireless access set to force a 'great IoT re-alignment' in 2026

• LoRa Alliance reports 125 million LoRaWAN end devices deployed globally

• Milesight FES Plus smart building solutions elevating buildings for greater intelligence and sustainability

🎙️ The IoT For All Podcast

In this episode of the IoT For All Podcast, Maor Efrati, co-founder and CTO of Monogoto, joins Ryan Chacon to discuss hybrid connectivity in IoT. The conversation covers why hybrid connectivity matters, the evolving landscape of IoT connectivity, the importance of edge AI and LPWAN, the critical role of built-in security in IoT solutions, the need for market education to dispel misconceptions about IoT connectivity, the impact of SIM virtualization on device management, hybrid connectivity with GEO and LEO satellies, and the future of hybrid connectivity.

YouTube • Spotify • Apple • YT Music • Amazon

🗓️ Events & Webinars

📄 White Papers