Industrial IoT Statistics 2026
By Axis Intelligence Research
Co-author: Sophie Winslow (Industrial Technology) | Last updated: August 27, 2026 | License: CC BY 4.0
Ericsson counts 22.3 billion IoT connections worldwide at the end of 2025, rising to 47.1 billion by 2031. Only 4.5 billion of those are cellular. Axis Intelligence Research finds that cellular’s share of the total connection base is set to shrink from 20.2% to 16.6% over that period, even as cellular IoT itself grows at 10% a year.
Quick Answer: How Big Is Industrial IoT in 2026?
There were 22.3 billion IoT connections globally at end-2025, of which 4.5 billion were cellular and 2.6 billion were Broadband or Critical IoT on 4G/5G, according to the June 2026 Ericsson Mobility Report. Connectivity is not the constraint. Deployment depth is: Axis Intelligence Research measures an Industrial Depth Ratio (IDR) of 0.544 for large EU enterprises against 0.247 for small ones — a 2.20× gap in how far IoT reaches past the security camera and into the production line.
Key Findings
- Axis Intelligence Research finds that small and large EU enterprises deploy IoT for premises security at an identical rate — 72% of IoT users in both size classes — while large enterprises run 2.20× the Industrial Depth Ratio, the gap sitting entirely in production-process and condition-based-maintenance use.
- Axis Intelligence Research calculates that 17.28% of large EU enterprises use IoT in a production process, against 3.90% of small ones — a 4.43× spread derived from Eurostat adoption and purpose data.
- According to the Ericsson Mobility Report published in June 2026, total cellular IoT connections reached around 4.5 billion at the end of 2025 and are forecast to approach 7.8 billion by 2031.
- Axis Intelligence Research calculates that the per-sector values in the FBI’s 2025 critical-infrastructure chart sum to 2,118 ransomware complaints, or 58.7% of all 3,611 US ransomware complaints recorded that year by the Internet Crime Complaint Center.
- The International Federation of Robotics recorded 542,000 industrial robot installations in 2024 and a global operational stock of 4,664,000 units, a 9% year-on-year increase.
How Many Industrial IoT Devices Are Connected in 2026?
The honest answer is that nobody counts “industrial IoT devices” as a category, because no statistical agency defines one. What exists are connection counts by bearer technology, and enterprise survey data on what firms actually do with sensors. Both are published. Neither is a market-size guess.
Ericsson’s June 2026 forecast is the cleanest connection-side series available, and it breaks the base into three buckets.
Global IoT Connections by Segment, 2025 and 2031
| Segment | End-2025 | End-2031 (forecast) | CAGR | Source |
|---|---|---|---|---|
| Short-range IoT | 17.5bn | 38.8bn | 14% | Ericsson Mobility Report, June 2026 |
| Wide-area IoT | 4.9bn | 8.3bn | 9% | Ericsson Mobility Report, June 2026 |
| — of which cellular IoT | 4.5bn | 7.8bn | 10% | Ericsson Mobility Report, June 2026 |
| Total | 22.3bn | 47.1bn | 13% | Ericsson Mobility Report, June 2026 |
Cellular figures are contained within wide-area IoT; the total does not double-count them.
Two things fall out of that table once you do the division. Axis Intelligence Research calculates that short-range technologies account for 85.9% of the 24.8 billion net connections added between 2025 and 2031 — (38.8 − 17.5) ÷ (47.1 − 22.3) — while cellular contributes 13.3%. And cellular’s share of the whole base falls from 20.2% to 16.6% despite growing by 3.3 billion connections.
Sophie Winslow, Industrial Technology: Measured against what baseline? That is the question the connection-count headlines never answer. A plant running a dense short-range mesh inside one building and a utility running a handful of cellular endpoints across dozens of pumping stations both show up in these tables, and they are not the same business. The 14% short-range growth rate is the number to watch for cycle-time and uptime work, because that is where the dense in-cell instrumentation lives. Cellular growth is a geography story — assets you cannot reach with a cable.
The Cellular Split: Massive IoT vs Broadband and Critical IoT
Within the 4.5 billion cellular connections, Ericsson puts Broadband and Critical IoT (4G/5G) at around 2.6 billion at end-2025, which Axis Intelligence Research computes as 57.8% of the cellular base, leaving 42.2% on Massive IoT technologies — NB-IoT and Cat-M. Ericsson reports that 240 million Massive IoT connections were added globally during 2025 alone.
Regional concentration is extreme. Ericsson states that North East Asia holds around 70% of all cellular IoT connections at the end of 2025. Any global cellular IoT statistic is therefore, to a first approximation, a statistic about one region’s smart-meter and utility rollouts.
5G RedCap Deployment Status
| Indicator | Value | As of | Source |
|---|---|---|---|
| Service providers with commercial 5G RedCap | 14 | June 2026 | Ericsson Mobility Report (citing GSA) |
| Service providers investing in 5G RedCap | 42 | June 2026 | Ericsson Mobility Report (citing GSA) |
| Countries with RedCap investment | 27 | June 2026 | Ericsson Mobility Report (citing GSA) |
| Service providers that finished 3G shutdown | 80 | End-2025 | Ericsson Mobility Report (citing GSA) |
| Service providers that finished 2G shutdown | 46 | End-2025 | Ericsson Mobility Report (citing GSA) |
Fourteen commercial RedCap launches is a small number for a technology that has been positioned as the 5G answer for mid-tier industrial endpoints since 2023. Ericsson expects enhanced RedCap — cheaper, capped at 10 Mbps peak — to reach full commercial potential in 2028 and beyond. For anyone planning a refresh cycle on cellular-connected machinery, that is the date that matters, not the RedCap launch count.
The sunset numbers carry a harder operational edge. Eighty service providers have completely finished shutting down 3G, and 46 have fully turned off 2G. Every one of those events strands legacy industrial modems that were specified a decade ago on the assumption that 2G would outlive the machine.
Industrial IoT Adoption Rates: What Enterprises Actually Do With Sensors
Connection counts describe supply. Adoption surveys describe use, and the only official EU-wide measurement that separates IoT purposes comes from Eurostat’s ICT usage in enterprises survey, dataset isoc_eb_iot. The purpose module was fielded in the 2021 wave across a sample of 148,000 enterprises drawn from roughly 1.5 million EU firms with ten or more people employed. [older data] — no newer EU-wide wave has published the purpose breakdown, and the size-class contrast it captures is structural rather than cyclical, which is why it remains the reference series.
EU Enterprise IoT Adoption by Size Class
| Size class | Using IoT | Source |
|---|---|---|
| Large (250+) | 48% | Eurostat, isoc_eb_iot, 2021 |
| Medium (50–249) | 37% | Eurostat, isoc_eb_iot, 2021 |
| Small (10–49) | 26% | Eurostat, isoc_eb_iot, 2021 |
| All enterprises 10+ | 29% | Eurostat, isoc_eb_iot, 2021 |
Purpose of IoT Deployment (% of enterprises using IoT)
| Purpose | Large | Small | All sizes | Source |
|---|---|---|---|---|
| Premises security | 72% | 72% | 72% | Eurostat, isoc_eb_iot, 2021 |
| Energy management | 56% | 26% | 30% | Eurostat, isoc_eb_iot, 2021 |
| Condition-based maintenance | 44% | 22% | 24% | Eurostat, isoc_eb_iot, 2021 |
| Production process | 36% | 15% | — | Eurostat, isoc_eb_iot, 2021 |
That premises-security row is the most under-reported line in enterprise IoT data. A small workshop and a large manufacturer secure their sites with connected cameras at the same rate. Every other purpose diverges by a factor of two or more.
The Axis Industrial Depth Ratio (IDR)
IDR — Industrial Depth Ratio. A single reading of how far a population’s IoT deployment reaches past facility management and into the production asset base.
Formula, weights, and inputs — reproducible from published Eurostat values:
IDR = (0.6 × P + 0.4 × M) ÷ S
P = production-process use, % of enterprises using IoT
M = condition-based-maintenance use, % of enterprises using IoT
S = premises-security use, % of enterprises using IoT
Production-process integration carries the heavier weight (0.6) because it implies live coupling between sensing and the control layer — MES, SCADA, or PLC-adjacent systems — whereas condition-based maintenance (0.4) can and often does run as a read-only overlay on existing machinery. Premises security is the denominator because it is the universal entry-level deployment: the thing almost every IoT-using firm does first, and therefore the fairest baseline against which to measure everything that comes after.
IDR Readings, EU Enterprises
| Population | P | M | S | Numerator | IDR | Source |
|---|---|---|---|---|---|---|
| Large enterprises (250+) | 36% | 44% | 72% | 39.2 | 0.544 | Axis Intelligence Research, from Eurostat isoc_eb_iot |
| Small enterprises (10–49) | 15% | 22% | 72% | 17.8 | 0.247 | Axis Intelligence Research, from Eurostat isoc_eb_iot |
| Gap | — | — | — | — | 2.20× | Axis Intelligence Research |
Working: large = (0.6 × 36 + 0.4 × 44) ÷ 72 = 39.2 ÷ 72 = 0.544. Small = (0.6 × 15 + 0.4 × 22) ÷ 72 = 17.8 ÷ 72 = 0.247. Gap = 0.544 ÷ 0.247 = 2.20.
The denominator is identical in both rows, which is the finding. The industrial-depth gap between large and small European firms is not a gap in whether they adopt IoT, and not a gap in how they secure their buildings. It sits entirely in the numerator — in whether the sensors ever touch a production asset.
Run the same inputs against the whole enterprise base rather than against IoT users, and the spread widens. Axis Intelligence Research calculates production-purpose penetration as adoption rate × purpose share:
- Large: 48% × 36% = 17.28% of all large EU enterprises
- Small: 26% × 15% = 3.90% of all small EU enterprises
- Ratio: 4.43×
Condition-based maintenance runs 21.12% against 5.72% on the same method. Premises security runs 34.56% against 18.72% — a gap of under two-fold.
Interpretation, in the silo author’s voice — Sophie Winslow: Two readings, one denominator, and the whole Industry 4.0 adoption debate reorganises itself. Small manufacturers are not sensor-averse; they buy cameras at exactly the rate the big plants do. What they do not buy is the integration work. A condition-based maintenance programme needs a baseline, a shift pattern to measure against, and somebody who owns the alert at 3am. That is headcount, not hardware, and it is the reason IDR moves with firm size while the security line does not. What would move the reading: managed condition-monitoring sold as a service rather than as a capital project. Watch the numerator, not the adoption rate.
Scope of the metric. IDR is computed from a multi-response survey question, so its components overlap and it is a ratio of penetrations rather than a share of deployments. It reads the EU enterprise population as measured in the 2021 wave and does not extend to firms with fewer than ten employees, which the survey covers only optionally. It is not a maturity score for an individual plant.
Industrial IoT Adoption by Sector
Sector-level readings from the same Eurostat wave show where sensing has actually landed.
IoT Adoption Rate by Economic Activity
| Sector | Using IoT | Source |
|---|---|---|
| Electricity, gas, steam, air conditioning, water supply | 46% | Eurostat, isoc_eb_iot, 2021 |
| Accommodation | 36% | Eurostat, isoc_eb_iot, 2021 |
| Information and communication | 34% | Eurostat, isoc_eb_iot, 2021 |
| Professional, scientific and technical | 25% | Eurostat, isoc_eb_iot, 2021 |
| Administrative and support services | 24% | Eurostat, isoc_eb_iot, 2021 |
Purpose Concentration by Sector (% of that sector’s IoT users)
| Reading | Sector | Value | Source |
|---|---|---|---|
| Production-process use, highest | Electricity, gas, steam, water | 34% | Eurostat, isoc_eb_iot, 2021 |
| Production-process use, manufacturing | Manufacturing | 32% | Eurostat, isoc_eb_iot, 2021 |
| Condition-based maintenance, highest | Electricity, gas, steam, water | 41% | Eurostat, isoc_eb_iot, 2021 |
| Logistics use, highest | Transportation and storage | 61% | Eurostat, isoc_eb_iot, 2021 |
| Premises security, highest | Retail trade | 79% | Eurostat, isoc_eb_iot, 2021 |
| Premises security, lowest | Transportation and storage | 61% | Eurostat, isoc_eb_iot, 2021 |
Utilities out-index manufacturing on both production-process and condition-based-maintenance use. That runs against the way industrial IoT is usually marketed — as a factory-floor story — and it has a straightforward explanation: a distribution network is a set of remote, unstaffed, revenue-critical assets whose failure is expensive and whose inspection is manual. That is the highest-return sensing case in the economy, and the sector found it first.
Industrial IoT Adoption by Country
| Country | Enterprises using IoT | Source |
|---|---|---|
| Austria | 51% | Eurostat, isoc_eb_iot, 2021 |
| Slovenia | 49% | Eurostat, isoc_eb_iot, 2021 |
| Finland | 40% | Eurostat, isoc_eb_iot, 2021 |
| Sweden | 40% | Eurostat, isoc_eb_iot, 2021 |
| Poland | 19% | Eurostat, isoc_eb_iot, 2021 |
| Estonia | 17% | Eurostat, isoc_eb_iot, 2021 |
| Bulgaria | 15% | Eurostat, isoc_eb_iot, 2021 |
| Romania | 11% | Eurostat, isoc_eb_iot, 2021 |
Forty percentage points separate Austria from Romania inside a single market with a single set of product rules. Whatever the EU-wide 29% average describes, it is not a condition any member state is actually in.
Industrial Automation Statistics: The Installed Base Sensors Attach To
Sensing does not deploy into empty buildings. It deploys onto machinery, and the machinery count is published annually by the International Federation of Robotics.
Global Industrial Robot Installations and Stock
| Indicator | Value | Period | Source |
|---|---|---|---|
| Annual installations | 542,000 units | 2024 | IFR, World Robotics 2025 |
| Operational stock | 4,664,000 units | 2024 | IFR, World Robotics 2025 |
| Stock growth | 9% | 2023→2024 | IFR, World Robotics 2025 |
| Installations forecast | 575,000 units | 2025 | IFR, World Robotics 2025 |
| Installations forecast | >700,000 units | 2028 | IFR, World Robotics 2025 |
Installations by Market, 2024
| Market | Units | Change | Source |
|---|---|---|---|
| China | 295,000 | Record high | IFR, World Robotics 2025 |
| Japan | 44,500 | −4% | IFR, World Robotics 2025 |
| United States | 34,200 | −9% | IFR, World Robotics 2025 |
| Republic of Korea | 30,600 | −3% | IFR, World Robotics 2025 |
| Germany | 26,982 | −5% | IFR, World Robotics 2025 |
| India | 9,100 | +7% | IFR, World Robotics 2025 |
| Italy | 8,783 | −16% | IFR, World Robotics 2025 |
| Mexico | 5,600 | −4% | IFR, World Robotics 2025 |
| Spain | 5,100 | — | IFR, World Robotics 2025 |
| France | 4,900 | −24% | IFR, World Robotics 2025 |
| Canada | 3,800 | −12% | IFR, World Robotics 2025 |
| United Kingdom | 2,500 | −35% | IFR, World Robotics 2025 |
Regional shares in 2024: Asia 74%, Europe 16%, the Americas 9%. China’s operational stock reached 2,027,000 units, which Axis Intelligence Research calculates as 43.5% of the global installed base. Chinese suppliers took 57% of their home market in 2024, up from around 28% over the past decade — the first year domestic vendors outsold foreign ones in China.
Europe installed 85,000 units, down 8%, with 67,800 of those inside the EU. The Americas installed 50,100, down 10%, of which the United States took 68%.
Sophie Winslow: Eight of the twelve largest markets shrank in 2024, and IFR still forecasts 575,000 units for 2025 — a 6.1% rebound on the published 542,000 base. That forecast is doing a lot of work. The UK’s 35% drop is not a demand signal; it is the hangover from a super-deduction tax credit that expired in Q1 2023 and pulled a year of orders forward. Read national robot numbers with the incentive calendar open next to them, or you will call a policy artefact a trend.
Industrial IoT Security Statistics: What the Exposure Actually Looks Like
Connected production assets create a reporting trail, and in the United States that trail runs through the FBI’s Internet Crime Complaint Center. The 2025 IC3 Annual Report recorded 3,611 ransomware complaints with $32,320,105 in reported losses — a figure that excludes downtime, equipment, and third-party remediation, which is where the real cost of an OT incident sits.
US Ransomware Reporting, Three-Year Series
| Year | Ransomware complaints | Reported losses | Source |
|---|---|---|---|
| 2023 | 2,825 | $59,641,384 | FBI IC3 Annual Report |
| 2024 | 3,156 | $12,473,156 | FBI IC3 Annual Report |
| 2025 | 3,611 | $32,320,105 | FBI IC3 Annual Report |
Axis Intelligence Research sums the per-sector values published in the report’s critical-infrastructure chart to 2,118 ransomware complaints and 1,119 data-breach complaints across the 16 critical infrastructure sectors — meaning 58.7% of all US ransomware complaints in 2025 came from critical infrastructure (2,118 ÷ 3,611). The single highest sector reading was 460 ransomware complaints and 182 data breaches, in healthcare and public health, which Axis Intelligence Research calculates as 21.7% of the critical-infrastructure ransomware total.
IC3 names Critical Manufacturing among the three sectors most affected by the ten most-reported ransomware variants, alongside healthcare and government facilities. Those ten variants accounted for 56.8% of reported incidents and 49.8% of reported losses. IC3 identified 63 new variants during the year, an average of 5.25 a month.
Beyond the critical-infrastructure sectors, IC3 logged more than 1,400 ransomware complaints from other organisations, with non-critical manufacturing — furniture and building materials, in the report’s own examples — accounting for 5% of them.
Recovery data tells a separate story. The IC3 Recovery Asset Team initiated 655 Financial Fraud Kill Chain actions for critical-infrastructure entities in 2025, freezing $146,561,094 of $261,451,001 in reported losses, a 56% success rate.
Marcus Chen, Cybersecurity & Privacy: No zero-day required. The variants at the top of that list — Akira, Qilin, INC/Lynx/Sinobi — run on compromised credentials, then delete backups and encrypt. The reason Critical Manufacturing keeps appearing in these tables is not that PLCs are uniquely fragile; it is that a line stoppage prices downtime by the hour, and attackers read the same operations research everybody else does. Compare 2025’s 3,611 complaints with 2023’s 2,825 and the direction is clear, but treat the loss column with suspicion in both years: a $12.5m total for 2024 against $59.6m for 2023 reflects who filled in the loss field, not what the year cost.
An overlooked signal sits in the report’s own appendix. Axis Intelligence Research counted four OT-specific joint guidance products in IC3’s 2025 industry alert log: Secure by Demand for OT owners and operators (January 13), Primary Mitigations to Reduce Cyber Threats to Operational Technology (May 6), Foundations for OT Cybersecurity asset inventory guidance (August 13), and guidance on maintaining a definitive view of OT architecture (September 29). Three of the four are about knowing what you have. That is the state of OT asset inventory in 2026, stated by the agencies rather than by a vendor survey.
What Changes for Industrial IoT Vendors on 11 September 2026?
Fifteen days from this update, the EU Cyber Resilience Act’s reporting obligations take effect. The European Commission confirms that from 11 September 2026, manufacturers of products with digital elements must notify actively exploited vulnerabilities and severe incidents.
CRA Reporting Timeline
| Obligation | Deadline | Source |
|---|---|---|
| Early warning | 24 hours from awareness | European Commission |
| Full notification | 72 hours | European Commission |
| Final report (exploited vulnerability) | 14 days after a corrective measure is available | European Commission |
| Final report (severe incident) | Within one month | European Commission |
Reports route once, through the CRA Single Reporting Platform, to the CSIRT where the manufacturer has its main establishment and simultaneously to ENISA. The Commission states the platform will be operational by 11 September 2026, with functional and security testing under way.
The date most industrial vendors have written down is 11 December 2027, when the CRA’s main obligations apply — 471 days out from this update. The reporting duty arrives first, and it applies to the installed base, not just to new catalogue items. For a sector whose products carry ten- to twenty-year service lives, that ordering is the whole story.
Marcus Chen: A 24-hour clock is an operational spec, not a compliance line item. It means somebody has to be reachable, somebody has to be able to say which shipped firmware versions contain the affected component, and somebody has to be authorised to file before legal has finished reading. Vendors who cannot answer the SBOM question in an afternoon today will not answer it in 24 hours in September.
Methodology
Collection. Every figure in this report was retrieved from a primary source document during production on 27 August 2026 and logged with its URL and retrieval date in the accompanying CSV. Six source organisations were used: Ericsson (Mobility Report, June 2026 edition), Eurostat (isoc_eb_iot, ICT usage in enterprises), the FBI Internet Crime Complaint Center (2025 IC3 Annual Report), the International Federation of Robotics (World Robotics 2025), and the European Commission (Cyber Resilience Act reporting obligations). No market-sizing estimate from a commercial research vendor appears anywhere in this dataset.
Arithmetic. All Axis-calculated values were computed twice independently — once in floating point and once in exact rational arithmetic — and both passes agreed to nine decimal places before any number was written into prose.
Formulas disclosed in full.
IDR = (0.6 × production_process + 0.4 × condition_based_maintenance) ÷ premises_security, all components expressed as % of enterprises using IoT, from a single survey wave.- Purpose penetration of the full enterprise base =
adoption_rate × purpose_share_among_users. Valid because Eurostat expresses purpose figures as a share of IoT-using enterprises. - Critical-infrastructure ransomware total = arithmetic sum of the per-sector values printed in the IC3 2025 critical-infrastructure chart.
- Segment shares of the IoT connection base =
segment ÷ total, using Ericsson’s own rounded published figures.
Scope. Eurostat’s purpose module reports the 2021 survey wave and is flagged [older data] in the CSV; it remains the only official EU-wide measurement separating industrial from facility IoT purposes. Eurostat purpose questions are multi-response, so component values overlap and cannot be summed to 100%. Ericsson connection figures are rounded by the publisher, so derived shares carry rounding at the first decimal. IC3 figures capture complaints filed with one US agency and undercount both incidents and losses by the agency’s own statement. IFR counts industrial robots, not the sensing layer attached to them; it is used here as the installed-base denominator, not as an IoT measure.
Disagreement on the record. IFR’s 575,000-unit forecast for 2025 implies a 6.1% rebound after a year in which eight of the twelve largest national markets contracted. Axis Intelligence Research treats that forecast as the upper bound rather than the central case, on the basis that the 2024 declines in France (−24%) and the UK (−35%) were driven by incentive expiry and investment-cycle timing, neither of which reverses on the forecast horizon.
About This Dataset
Title: Industrial IoT Statistics 2026 — Connections, Enterprise Adoption Depth, Automation Installed Base, and OT Exposure
Creator and publisher: Axis Intelligence Research
Temporal coverage: 2021–2031 (observations 2021–2025; forecasts to 2031)
Spatial coverage: Global, European Union and member states, United States, major robotics markets
Rows: 140 observations, each carrying source organisation, source document, source URL, retrieval date, primary-source flag, and — where Axis-calculated — the full formula in method_note
Format: UTF-8 CSV, one header row, long format, ISO 8601 dates, raw values with units in a separate column
File: industrial-iot-statistics-2026.csv
License: Creative Commons Attribution 4.0 International (CC BY 4.0). Free to share and adapt, including commercially, with attribution.
Cite This Research
APA: Axis Intelligence Research. (2026, August 27). Industrial IoT statistics 2026: Connections, adoption depth, and OT exposure. Axis Intelligence. https://axis-intelligence.com/industrial-iot-statistics/
MLA: Axis Intelligence Research. “Industrial IoT Statistics 2026: Connections, Adoption Depth, and OT Exposure.” Axis Intelligence, 27 Aug. 2026, axis-intelligence.com/industrial-iot-statistics/.
Chicago: Axis Intelligence Research. “Industrial IoT Statistics 2026: Connections, Adoption Depth, and OT Exposure.” Axis Intelligence, August 27, 2026. https://axis-intelligence.com/industrial-iot-statistics/.
Copy any format above. The underlying CSV is CC BY 4.0 and may be redistributed with attribution.
Frequently Asked Questions
Why does the Industrial Depth Ratio use premises security as its denominator instead of total headcount or revenue?
Because premises security is the one IoT purpose deployed at an identical rate by large and small enterprises — 72% of IoT users in both size classes. That makes it a natural control: it removes firm-size effects on the willingness to connect anything at all, and isolates the variable that actually differs, which is whether sensing reaches production assets. A headcount or revenue denominator would re-import the size effect the metric is designed to strip out.
Is condition-based maintenance the same thing as predictive maintenance in these statistics?
No, and the distinction matters when reading any adoption figure. Eurostat measures sensors monitoring the real-time condition of equipment. That covers vibration, temperature, and current-draw monitoring whether or not any model sits on top of the data. Predictive maintenance is a modelling layer built on the same feed. A firm answering yes to Eurostat’s question may be doing threshold alerting and nothing more.
Why do utilities out-index manufacturing on production-process IoT use?
Utilities report 34% production-process use among IoT adopters against manufacturing’s 32%, and 41% on condition-based maintenance — the highest sector reading. Distribution networks consist of remote, unstaffed assets whose failure is immediately revenue-affecting and whose manual inspection requires a truck roll. That combination produces a shorter payback than an in-plant sensor does, because a factory already has staff walking past the machine.
Does 5G RedCap change the economics of connecting mid-tier industrial equipment?
Not yet at scale. Fourteen service providers had commercially launched RedCap as of the June 2026 Ericsson data, with 42 investing across 27 countries. Ericsson positions enhanced RedCap — lower cost, 10 Mbps peak, comparable to LTE Cat-1 — as reaching full commercial potential in 2028 and beyond. Equipment specified for a refresh before then is being specified into an LTE Cat-1 world, not a RedCap one.
What happens to industrial equipment still running on 2G or 3G modems?
It stops reporting when the carrier switches off the band. Eighty service providers had completely finished 3G shutdown by the end of 2025, and 46 had fully turned off 2G. Ericsson expects 3G sunsets to continue faster than 2G, because refarming 3G spectrum strengthens 4G and 5G, while 2G uses minimal spectrum and continues to carry legacy IoT traffic. Asset owners with fleet-wide cellular telemetry should be auditing modem generations against their operator’s published sunset schedule, not against a global average.
How much of US ransomware activity actually reaches operational technology rather than corporate IT?
The published data does not separate them, and any statistic claiming to is inferring. What IC3 does report is that critical-infrastructure sectors account for 2,118 of 3,611 ransomware complaints by Axis Intelligence Research’s summation of its sector chart, and that Critical Manufacturing sits among the three sectors most affected by the top ten variants. Those variants enter through compromised credentials and corporate systems; whether they cross into the control network is an incident-by-incident question that complaint data cannot answer.
Does the Cyber Resilience Act apply to industrial equipment already installed in EU plants?
The reporting obligation does. From 11 September 2026, manufacturers must notify actively exploited vulnerabilities and severe incidents in products with digital elements, and the European Commission’s guidance is explicit that reporting is the first obligation to apply — ahead of the main requirements on 11 December 2027. For equipment with a fifteen-year service life, the practical question is whether the vendor can identify affected firmware versions inside 24 hours.
Why is cellular IoT’s share of total connections falling while cellular IoT grows?
Because short-range connections are growing faster. Ericsson’s forecast has short-range IoT rising from 17.5bn to 38.8bn between 2025 and 2031 at a 14% CAGR, against cellular’s 10%. Axis Intelligence Research calculates that short-range accounts for 85.9% of net connection additions over the period and cellular for 13.3%. Cellular remains the bearer for geographically distributed assets; short-range is where dense in-facility instrumentation is going.
Which country-level IoT adoption figure should be used for European benchmarking?
Not the EU average. Austria reports 51% and Romania 11% — a 40-percentage-point spread inside a single regulatory market. Benchmarking a Polish plant (19%) against the 29% EU figure produces a misleading gap, and benchmarking an Austrian one against it produces false comfort. Use the member-state reading and state the survey wave.