Industrial networking
Industrial network market share: Ethernet, PROFINET, EtherNet/IP, EtherCAT
Industrial Ethernet now accounts for 79% of newly installed industrial network nodes, up from 34% when the count began in 2015. Fieldbus, which once carried most new installations, is down to 14%.
The individual protocol shares are where citations go wrong. Through 2024 they were reported as shares of the whole market: PROFINET 23%, EtherNet/IP 21%, EtherCAT 16%. The 2026 study reports them within Industrial Ethernet instead, where PROFINET is 30%. Those two 'PROFINET' numbers do not describe the same quantity.
This page assembles the year-by-year series in one place, keeps the two denominators apart, and shows what the protocol trend actually looks like once they are.
Data covers Newly installed industrial network nodes, 2017 to 2026 (HMS Networks annual study). Edited by Mike Ramsey / Reliable Media.
The figures and where they come from
Each figure is rated for how safely you can cite it today. Ratings judge current usability, not whether a number was ever correct.
| Figure | What it is | Source | Citation Confidence | Notes |
|---|---|---|---|---|
| 79% of new nodes | Industrial Ethernet share, 2026 | [A] | High | The current figure. Ethernet's share of newly installed nodes, not of the installed base, which still contains a great deal of fieldbus. |
| 71% | Industrial Ethernet share, 2024 | [B] | High | Two years earlier. The series moves several points a year, so the year matters more than the number. |
| 66% | Industrial Ethernet share, 2022 | [C] | High | The 2022 study. |
| 64% | Industrial Ethernet share, 2020 | [D] | High | The 2020 study, when fieldbus was still 30%. |
| 46% | Industrial Ethernet share, 2017 | [E] | High | The 2017 study, the year Ethernet was still under half of new nodes. |
| 34% | Industrial Ethernet share, 2015 | [A] | Medium | The starting point, quoted retrospectively in the 2026 release rather than measured in a release we cite directly. |
| 30% of Ethernet | PROFINET, 2026 (share of Ethernet) | [A] | High | Reported by the 2026 study as a share WITHIN Industrial Ethernet, not of the whole market. Against Ethernet's 79% total share this is roughly 24% of all new nodes, close to the 23% reported for 2024 on the old basis. Do not read it as a jump from 23% to 30%. |
| 25% of Ethernet | EtherNet/IP, 2026 (share of Ethernet) | [A] | High | Also a within-Ethernet share. Second behind PROFINET on that basis. |
| 20% of Ethernet | EtherCAT, 2026 (share of Ethernet) | [A] | High | A within-Ethernet share, third behind PROFINET and EtherNet/IP. Not comparable with the 16% whole-market figure reported for 2024. |
| 14% | Fieldbus share, 2026 | [A] | High | Fieldbus as a whole, down from 17% a year earlier. Still a large installed base, but a shrinking share of what gets installed new. |
| 23% | PROFINET, 2024 | [B] | High | PROFINET took the outright lead in 2024 after years of near-parity. Reported as a share of the WHOLE market, which is the basis used through 2024. |
| 21% | EtherNet/IP, 2024 | [B] | High | The 2024 figure. |
| 16% | EtherCAT, 2024 | [B] | High | Third in 2024, on the whole-market basis. |
Why the numbers disagree
The trap on this topic is the denominator, and it is recent. Through the 2024 study, protocol shares were reported against the whole market for newly installed nodes: PROFINET 23%, EtherNet/IP 21%, EtherCAT 16%. The 2026 study reports them within Industrial Ethernet instead. Set 23% next to 30% and PROFINET looks like it gained seven points. Convert the 2026 figure back, 30% of a 79% Ethernet share, and it lands near 24% of all new nodes, which is roughly flat.
The same correction applies to every protocol in the 2026 list. EtherCAT at 20% within Ethernet is close to 16% of the whole market, the figure reported for 2024. Any protocol trend drawn straight from headline numbers across 2024 and 2026 overstates growth, because the base shrank from the whole market to a 79% slice of it.
The Ethernet and fieldbus totals do not have this problem. Both are reported against all newly installed nodes in every year, which is why the 34% to 79% Ethernet series is the one part of this data that can be quoted as a clean trend.
It is also one company's estimate. HMS Networks is a supplier of industrial communication hardware and publishes this study annually from its own market view. No competing public series exists at this granularity, which is exactly why it is universally cited. Treat it as the best available estimate, not an audited census.
How to cite these figures
For the headline trend, use the Ethernet share of all new nodes: 34% in 2015, 46% in 2017, 64% in 2020, 66% in 2022, 71% in 2024, 79% in 2026. Consistent denominator, clean trend.
For protocol shares, state the basis. Either 'PROFINET held 23% of all newly installed nodes in the 2024 study' or 'PROFINET held 30% within Industrial Ethernet in the 2026 study.' Never mix the two in one sentence.
Say 'newly installed nodes' every time. None of these figures describe the installed base, which still contains far more legacy fieldbus than the new-node shares suggest.
Attribute it to HMS Networks and give the study year. It is a vendor-published estimate, and naming the source lets a reader weigh it.
Where people go wrong
Comparing the 2026 protocol shares with earlier years. 30% within Industrial Ethernet is roughly 24% of all new nodes, not a seven-point gain over 2024's 23%.
Reading these as installed-base shares. They count new nodes only, so they understate how much fieldbus is still running in service.
Quoting a protocol share undated. The ordering changed more than once, and the reporting basis changed too, so an undated figure is likely wrong about both the level and the leader.
Presenting it as an independent census. It is one hardware vendor's annual estimate, with no comparable public series to check it against.
Assuming regional shares match global ones. The source notes EtherNet/IP leads in North America while PROFINET dominates in Central Europe, so a global share can invert in a specific market.
How we checked
Every figure comes from an HMS Networks annual market-share release, five separate years of them, each retrieved and confirmed to contain the figure quoted. Where a year is represented by trade-press coverage of the study rather than the HMS newsroom, the source list says so.
We assembled the series because the value of these numbers is the trajectory and nobody publishes the trajectory in one place. Each row is independently verifiable against its own year's release.
Checking the denominators is what changed this page. The 2026 release states its protocol shares are 'within Industrial Ethernet' while earlier releases give shares of the whole market, so the two cannot be put in one column. We label each protocol figure with its basis rather than converting the 2026 numbers, because a converted figure would be our arithmetic rather than the source's.
The 2015 starting figure is the one number quoted retrospectively, in the 2026 release, rather than taken from a 2015 release we cite. It is rated Medium for that reason.
We looked for a second, independent measurement of industrial protocol share and did not find one. Standards bodies publish node counts for their own protocol and nothing comparable across protocols, so there is no way to triangulate this series.
Full source list
Primary sources, with live links. Every figure above traces to one of these.
- [A]HMS Networks (via Industrial Ethernet Book)2026
HMS Networks industrial network market shares 2026, reported by Industrial Ethernet Book
https://industrialethernet.net/news/tech-updates/2026-industrial-network-market-shares-annual-report-by-hms-networks/ - [B]HMS NetworksJune 17, 2024
HMS Networks, "Annual analysis reveals steady growth in industrial network market" (2024 shares)
https://www.hms-networks.com/news/news-details/17-06-2024-annual-analysis-reveals-steady-growth-in-industrial-network-market - [C]HMS NetworksMay 1, 2022
HMS Networks, "Industrial networks keep growing despite challenging times" (2022 shares)
https://www.hms-networks.com/news/news-details/01-05-2022-industrial-networks-keep-growing-despite-challenging-times - [D]HMS Networks2020
HMS Networks, industrial network market shares 2020
https://www.hms-networks.com/news/news-details/-industrial-network-market-shares-2020-according-to-hms-networks - [E]HMS Networks2017
HMS Networks, industrial network market shares 2017
https://www.hms-networks.com/news/news-details/-industrial-ethernet-and-wireless-are-growing-fast---industrial-network-market-shares-2017-according-to-hms
Common questions
- What is the most used industrial network protocol?
- PROFINET, on both bases. It held 23% of all newly installed nodes in the 2024 study, and 30% within Industrial Ethernet in the 2026 study. EtherNet/IP is second on both.
- Did PROFINET jump from 23% to 30%?
- No, and this is the most common error with these figures. The 23% is a share of the whole market from the 2024 study; the 30% is a share within Industrial Ethernet from the 2026 study. Converted to the same basis, 30% of a 79% Ethernet share is roughly 24% of all new nodes, so PROFINET is close to flat.
- Is fieldbus dead?
- Not in service, but it is shrinking fast in new builds. Fieldbus fell to 14% of newly installed nodes in 2026, from 22% in 2024 and 30% in 2020. A great deal of PROFIBUS and DeviceNet is still running on existing plant floors.
- How much has Industrial Ethernet grown?
- From 34% of newly installed nodes in 2015 to 79% in 2026, passing 46% in 2017, 64% in 2020, 66% in 2022 and 71% in 2024. This is the one series here with a consistent denominator throughout.
- Who publishes these numbers?
- HMS Networks, a supplier of industrial communication hardware, in an annual study. Almost every published claim about protocol market share traces to it, and there is no independent public series at the same granularity.
More data, traced to source
- How many manufacturers use robots? Canada 8.4%, UK 14%, EU 16%
Three national statistics agencies measured how many manufacturers use robots and got 8.4%, 14% and 16%. All three are right. The gap is definitions and years, and it is why a single global adoption rate does not exist.
- How many US firms actually use robots? Far fewer than you would think
The story is that everyone is automating. The US Census Bureau's own firm survey found only about 2% of US firms use robotics. A newer Census release puts it far higher. The gap is the whole story.
- AI and automation adoption is a big-firm story: 37% at 250+ staff
The 'everyone is adopting AI' story is really a large-firm story. 37% of US firms with at least 250 employees use AI, against a much smaller share of small firms. The automation frontier is concentrated, not universal.