Hello readers!

This week, we’re talking about predictive maintenance finally catching up to legacy plant data, Low-Power Wide-Area Networks and eSIM Orchestration, and more!

A single vacuum pump failure in a semiconductor fab can scrap more than 100 wafers and cost a customer up to $2 million a year — a number Edwards, which builds much of the vacuum equipment chip fabs run on, publishes for its own installed base. The same shape shows up everywhere these pumps run. At STMicroelectronics' Catania fab, unplanned failures used to account for most pump replacements; five years into a predictive maintenance rollout, they're down to 10 percent, and the fab now goes 20,000 hours between preventive swaps instead of 10,000. Nanoprecise has documented a comparable shift at a North American fab running hundreds of dry pumps, where sensors caught early-stage wear in time to schedule a replacement instead of losing a batch. Three vendors, three fabs, one pattern: pumps that used to run on a fixed schedule or run to failure are now running on years of accumulated data.

SCADA and PLC systems have been logging plant data since the 1960s, built for a narrow job — local, closed-loop control, sensed and acted on in the moment. Push the sampling rate up high enough to train a predictive model, and a system built for that job starts to buckle: storage fills, throughput chokes, and the fab hits a ceiling that has nothing to do with the pumps themselves. The standard fix is a second, parallel data channel added alongside the original control system. The PLC keeps logging as it always has. A separate pipeline routes the same signals into a platform built to hold years of history instead of the next five minutes.

Maintenance already eats between 15 and 70 percent of a manufacturer's total cost of goods sold, split between two piles: parts and labor for failures nobody saw coming, and parts and labor for scheduled swaps the equipment didn't need yet. Plants that move to predictive, data-driven maintenance report close to 20 percent less downtime than those still running on fixed intervals. Put a number on either side of that gap — $2 million a year per pump failure, or a 50 percent cut in preventive replacements at Catania — and the case for building the parallel channel writes itself.

Every one of these rollouts frees up the same kind of person. Edwards' own engineers now spend their time reading pump behavior and building customer proposals, work that used to compete with hours lost untangling storage bottlenecks that had nothing to do with vacuum systems. Catania built a data-driven culture around the same shift, pairing its process engineers with outside data scientists instead of hiring a parallel infrastructure staff. Nanoprecise's customers get the same trade: sensor data lands in a platform already built to interpret it, so the maintenance team spends its hours acting on signals instead of building the pipeline that produces them.

None of this needed a new sensor or a breakthrough model. These fabs mostly built new models on data they were already generating — what changed was where that data could live and how far back it could see. The stress test is simple and doesn't require a vendor call to run: push the sampling rate on an existing SCADA or PLC system and watch where it breaks. That's the bottleneck, and it usually shows up years before anyone opens a model.

Longer-lasting battery technology is a lifesaver. Literally. If you were unlucky enough to need a pacemaker back in the 1950s, it needed to be wheeled around on a cart and plugged directly into a mains outlet. Your mobility was entirely dictated by the length of your extension cord.

Read more

Your business didn't plateau overnight. Growth slowed one bottleneck at a time.

But more hours at the office, more hires, and more marketing won't remove a hidden limitation. They just make it harder to spot.

The Revenue Ceiling Report reveals the four most common growth ceilings so you can identify what's quietly costing your business revenue and what to do next.

BELAY’s U.S.-based Assistants help build the systems and support that turn sustainable growth into reality.

Access the report for free

In this episode of the IoT For All Podcast, Wienke Giezeman, CEO and co-founder of The Things Industries, joins Ryan Chacon to discuss how IoT is finally delivering what it promised ten years ago. The conversation covers what changed technically and commercially, the ROI of IoT, why deployments failed in the early days, criticism of IoT, what companies still get wrong about LoRaWAN, and The Things Conference 2026.

YouTube • Spotify • Apple • YT Music • Amazon

Interested in becoming an IoT For All Partner? Reach out here!