Anybus ONE Device Intelligence: Devices That Learn from Their Own Data
HMS Networks has announced a "Device Intelligence" layer built into its Anybus ONE platform, designed to let industrial devices analyze their own operating data. This is a vendor announcement: here is what it says and what it means for those running plants.
On 29 October 2026, the magazine Automazione Plus published an interview with two HMS Networks executives: Bartek S. Candell, SVP Industrial Network Technologies, and Jörgen Palmhager, VP Technology in the same division. The focus is Anybus ONE Device Intelligence, an intelligence layer built into the new Anybus ONE embedded communication platform.
What was announced
Anybus solutions have long been known for connecting industrial devices to any network reliably and securely; according to HMS, millions of devices in the field rely on this technology. With Anybus ONE, the company adds a layer intended to let devices learn from their own operation, recognize changing conditions and turn the data they collect into better performance and greater reliability.
The starting problem
According to Candell, most industrial devices already measure quantities such as current, speed, pressure, temperature or cycle time. Until now, however, turning this data into useful information has required cloud platforms and data scientists. This will sound familiar to many companies: the signals are there, but getting from raw data to usable information remains costly and complex.
Why it matters for those running a plant
If analysis moves into the device and the connectivity layer, part of the processing happens at the edge, close to the machine. For plant managers and IT/OT leads, this suggests a scenario in which machine builders and component manufacturers could supply equipment able to flag changes in operating conditions on its own.
The connectivity question would then no longer be just "how do I get the data out?", but also "what information leaves the device already processed?". In the longer term, this could reduce the effort needed to turn machine signals into indicators usable by MES and monitoring systems, and it affects connectivity choices on new lines.
What we don't know yet
It should be remembered that the source is an interview with vendor executives, not an independent assessment: the claims are HMS's own. In addition, only the opening part of the source was reviewed, so information is missing on:
- the technical specifications of the solution;
- the artificial intelligence methods used;
- availability and pricing;
- customer examples or measured results.
It is therefore not possible, today, to assess how effective the approach really is.
A practical takeaway
Without anticipating results that have not been demonstrated, this announcement is worth using as a prompt for three checks in upcoming projects:
- ask machine and component suppliers which local analytics functions are planned, and what data they expose externally;
- define in advance which signals and events the MES and monitoring actually need, to avoid duplicating analysis between device and central systems;
- assess security, data governance and model updatability before relying on logic embedded in devices.
The value of on-board intelligence depends on what the device communicates and how the data integrates into plant systems.
Once technical details and use cases become available, it will be possible to tell whether Anybus ONE Device Intelligence is a concrete step toward more autonomous devices or just one more take on edge intelligence.