The IOT Express blog

Technical notes and perspectives on industrial IoT in the broad sense: connectivity and the CAN bus, error and state handling, multi-brand telematics, connected CMMS, data exploitation, artificial intelligence, digital sobriety… and everything that makes machines talk.

Home > Blog

2026

From ticket to spare part: a self-feeding knowledge base and online parts sales

When a problem gets solved through a discussion between a client and a technician, the solution often evaporates: it stays in one person’s head, or dissolves into an informal chat. By structuring each problem as a dedicated discussion, then summarising it with an LLM, that knowledge becomes a reusable, confidential base that no longer gets lost.

Read the article

2026

Industrial predictive maintenance: anticipating failure from the machine’s own data

On a nut harvester, a breakdown during the short harvesting window can compromise the entire season. Predictive maintenance does not guess the future: it continuously compares the machine’s real behaviour to its normal behaviour, and flags drift before the breakdown.

Read the article

2026

Non-intrusive energy monitoring: inferring a machine’s state without modifying it

Connecting each controller in a workshop individually (FANUC, SIEMENS, ABB and so on) is often impossible or too costly. Measuring electrical consumption and inferring the operating state from it bypasses the obstacle: no modification to the machine, yet a real-time view of the workshop.

Read the article

2026

Proving carbon savings with data: telematics and fleet decarbonisation

Claiming a cleaner fleet is no longer enough: customers and regulators ask for proof. By instrumenting the vehicles, you measure avoided consumption and carbon footprint with real data, and turn it into a verifiable commercial argument.

Read the article

2026

Real-time collection over intermittent connectivity: the quiet engineering of field data

A mobile machine works in a field, a tunnel, a warehouse: where the network is weak, intermittent and billed by volume. Collecting its data in real time without losing it, or blowing up the subscription cost, is an engineering problem in its own right.

Read the article

2026

From sensor to decision: the architecture of an end-to-end industrial IoT solution

Most industrial IoT players cover only one link of the chain. Here is why mastering the whole of it, from CAN frame decoding to artificial intelligence, is a game-changer for machine manufacturers.

Read the article

2026

Why multi-brand support changes everything in industrial telematics

Industrial fleets are rarely single-brand. Yet most telematics solutions are built for one specific manufacturer. Being agnostic, decoding any CAN frame, is a structural differentiator.

Read the article

2026

Digital sobriety as a product axis in industrial IoT

Collecting the full data of thousands of machines raises a volume problem. Treating the reduction of that volume as a product objective, rather than a constraint endured, serves cost, compliance and the environment at once.

Read the article