Industrial IoT (IIoT) connects your factory floor machines to business intelligence by attaching sensors to equipment, streaming that data through a connectivity layer, and feeding it into your ERP or BI dashboard so production decisions are based on live numbers rather than gut feel. For a UK SME factory, a sensible pilot covering 5 to 15 machines typically costs between £15,000 and £60,000, and well-run deployments cut unplanned downtime by 30 to 50% while lifting Overall Equipment Effectiveness (OEE) by 10 to 20 points. Made Smarter Adoption grants can fund 50% of eligible projects in supported English regions, and the PSTI Act 2022 plus NCSC operational technology guidance set the security baseline you must meet. The UK IoT manufacturing market is projected to reach around USD 8.95bn by 2026. Most factories see payback inside 9 to 18 months once downtime and energy savings land.
Last updated: June 2026
Industrial IoT is the practice of attaching network-connected sensors and controllers to physical production assets so the data they generate becomes visible, queryable and actionable in real time. The phrase that matters most is IT-OT convergence: tearing down the wall between Operational Technology (the PLCs, drives, robots and machines on the shop floor) and Information Technology (your ERP, CRM, BI tools and cloud). For decades these two worlds ran in parallel and rarely spoke. The machine knew it had stopped; the boardroom found out at the Monday meeting. IIoT closes that gap to seconds.
Here is why this matters commercially rather than technically. A UK manufacturer competing on margin cannot afford to discover scrap, downtime or energy waste a week after it happened. McKinsey research has repeatedly shown that connected operations can lift productivity by around 25%, cut maintenance costs by up to 40%, and reduce downtime by 30 to 50%. Those are not vendor fantasies; they are the difference between a profitable line and a loss-making one. The catch is the data-waste gap: McKinsey also found that 54% of companies use 10% or less of the IIoT data they collect. Connectivity alone is worthless. The value is in turning the stream into a decision.
Our view, after building these systems for UK firms, is that IIoT is badly oversold as a technology purchase and badly undersold as an operations programme. Bolting sensors onto machines is the easy 20%. The hard, valuable 80% is deciding which three numbers actually change behaviour on the shop floor, then making those numbers impossible to ignore. Be sceptical of any supplier who leads with hardware specifications before they have asked what decision you are trying to improve.
The building blocks of any IIoT system are consistent regardless of vendor:
Miss any one of those five and the project stalls. The most common failure we see in UK SMEs is investing heavily in the first four and forgetting the fifth, leaving a beautiful dashboard that nobody opens.
Factory floor data reaches a BI dashboard through a five-layer journey: sensor, edge, network, platform, and visualisation. Understanding this chain is the single most useful thing a non-technical manager can do, because it demystifies the whole project and lets you ask suppliers sharp questions at each stage. Let us walk the path of a single data point, say the spindle temperature on a CNC machine, from the metal to the boardroom screen.
First, a sensor on the spindle reads a temperature value. Second, an edge device or gateway sitting next to the machine collects that reading, often alongside dozens of others, and does light processing such as filtering noise or flagging an out-of-range value immediately so you do not wait for the cloud round trip. Third, the network layer moves the data, by wired Ethernet, industrial Wi-Fi, or in some UK sites private 5G, to wherever it is stored. Fourth, a platform (cloud services such as Azure IoT or AWS IoT, or an on-premise historian) stores the time-series data and runs analytics on it. Fifth, a visualisation and integration layer presents the result in a dashboard and pushes summarised figures into your ERP or MES so production planning reflects reality.
The table below maps the journey, the typical technology, and the question you should ask your integrator at each stage.
| Layer | What it does | Typical technology | Question to ask |
|---|---|---|---|
| Sensor | Reads the physical signal | Vibration, temperature, current, flow sensors | What accuracy and sample rate do I need? |
| Edge | Collects and pre-processes locally | Industrial gateways, edge PCs, IoT modules | Can it run if the internet drops? |
| Network | Moves data off the floor | Ethernet, industrial Wi-Fi, private 5G, LoRaWAN | Is OT traffic kept separate from office traffic? |
| Platform | Stores and analyses | Azure IoT, AWS IoT, on-prem historian | Do I own my data and can I export it? |
| Visualisation | Shows numbers and feeds ERP | Power BI, Grafana, MES/ERP connectors | Who acts on each metric, and when? |
The integration into ERP is where most of the business value lives and where most projects underinvest. When live machine output flows into your ERP, production planning stops being a spreadsheet guess. Stock reservations reflect actual throughput, delivery promises to customers reflect real capacity, and finance can see margin per job as it happens. If you already run an ERP, connecting IIoT data to it is often higher value than the sensors themselves; for firms considering a platform refresh, an well-planned Odoo ERP implementation in London can be designed from day one to ingest shop-floor data rather than bolting it on later.
The honest rule here: design the data flow backwards from the decision, not forwards from the sensor. Decide that the production planner needs live OEE per cell, then work back to the sensors required. Teams that start at the sensor end up drowning in data they never use, which is exactly the 54% waste figure in practice.
Yes, you can connect legacy machines that have no digital outputs, and for most UK factories this is the majority of the shop floor. The myth that you need to rip out and replace 20-year-old machinery to go digital is the single biggest reason owners delay IIoT, and it is wrong. Retrofitting is a mature, affordable discipline. A press from 1998 with no Ethernet port can still tell you when it is running, how fast, and whether it is about to fail, using bolt-on sensors that never touch the machine's control system.
There are three retrofit routes, in rising order of richness and cost:
The table below gives realistic UK cost bands per machine for each route in 2026, including hardware and integration labour.
| Retrofit route | Data richness | Cost per machine | Best for |
|---|---|---|---|
| Non-invasive sensing | Run state, cycles, vibration | £400 - £1,200 | Old machines, fast OEE pilots |
| Tapping existing signals | Run/idle/fault status | £600 - £1,800 | Machines with stack lights or relays |
| PLC integration | Full process data, alarms | £1,500 - £5,000 | Modern CNC, robots, packaging lines |
Our stance is firm: start every legacy factory with non-invasive sensing on the worst-performing cell. It is cheap, it cannot break the machine, and within a fortnight it produces a downtime chart that pays for the whole pilot in the conversation it triggers. We have watched a clamp-on sensor reveal that a "reliable" machine was actually idle 40% of the shift because of a material feed problem nobody had quantified. That single chart justified a six-figure programme. You do not need the expensive PLC integration on day one; earn it with results from the cheap stuff first.
Connecting the floor is also rarely the end goal in isolation. The data usually needs to drive workflows: a maintenance ticket raised automatically, a supervisor alerted by message, a re-order triggered. That orchestration layer is where business process automation in London turns raw signals into actions, so the system does something rather than just displaying a red number.
You should measure Overall Equipment Effectiveness (OEE) first because it compresses the three things that destroy manufacturing profit, availability, performance and quality, into one comparable score. OEE is the universal language of the shop floor. A line running at 100% OEE makes only good parts, as fast as it possibly can, with no stops. Most UK SME factories that have never measured it discover a true OEE between 45% and 65%, and they are genuinely shocked, because the number that lives in their head is closer to 85%. The gap between perceived and actual performance is where IIoT earns its keep.
OEE multiplies three factors:
The power of measuring these continuously, rather than on a clipboard once a shift, is that you stop arguing about anecdotes and start fixing causes. The biggest losses are usually the small, invisible ones: the two-minute stops that happen forty times a shift and never make it onto a fault log. Sensors catch every one.
The table below shows a realistic before-and-after for a single CNC cell after a six-month IIoT-driven OEE programme, using figures we consider typical for a UK SME.
| Metric | Before (manual) | After (IIoT) | Effect |
|---|---|---|---|
| Availability | 68% | 84% | Fewer micro-stops and faster changeovers |
| Performance | 76% | 88% | Slow running spotted and corrected |
| Quality | 94% | 98% | Drift caught before scrap is made |
| OEE | 49% | 72% | 23-point lift, more output, same machine |
Beyond OEE, the next metrics to wire up are energy consumption per unit, on-time delivery, and first-pass yield. Energy deserves special attention in the UK because industrial electricity prices have stayed punishingly high. Smart metering at machine level routinely reveals that a handful of assets, or worse, machines left running overnight doing nothing, drive a disproportionate share of the bill. We have seen factories cut energy spend 10 to 15% in the first quarter purely from visibility, before any process change.
The honest warning: do not put twenty metrics on a dashboard. Pick three that a supervisor can influence this week, make them visible on a screen on the floor, and review them daily. A focused dashboard that changes behaviour beats a comprehensive one that nobody reads. The 54% data-waste statistic is, fundamentally, a discipline problem, not a technology one.
Predictive maintenance pays for the whole project because unplanned downtime is the most expensive thing that happens in a factory, and predicting failure before it occurs converts emergency stoppages into scheduled, cheap interventions. The maths is brutal and in your favour. If a critical line costs you £1,200 an hour in lost output when it stops, and it suffers ten unplanned breakdowns a year averaging four hours each, that is 40 hours, or £48,000, vanishing annually from a single machine. Cut that by even 60% and you have saved nearly £29,000 a year, which is more than most pilots cost outright.
Predictive maintenance works by watching the early warning signs that humans cannot feel. Vibration sensors on motors, pumps and bearings detect the tiny imbalance that precedes a failure weeks ahead. Temperature sensors catch overheating before it seizes a component. Current draw signatures reveal a motor straining against a developing fault. The system learns each machine's healthy baseline, then flags deviation. Instead of a catastrophic 3am failure, you get a calm alert: bearing on pump 4 trending out of tolerance, schedule replacement within ten days.
Here is a worked ROI example for a single packaging line, the kind of one-page business case we build for UK clients.
| Item | Figure |
|---|---|
| Downtime cost per hour | £1,200 |
| Unplanned downtime before | 40 hours/year |
| Downtime reduction achieved | 60% |
| Annual downtime saving | £28,800 |
| Reduced maintenance and parts spend | £9,000 |
| Total annual benefit | £37,800 |
| Pilot cost (one line) | £22,000 |
| Payback period | Approximately 7 months |
The progression of maintenance maturity runs in four stages, and most UK SMEs sit firmly in the first two:
Our honest view: do not let a vendor sell you stage four AI before you have stage three condition monitoring working. The phrase "AI-powered predictive maintenance" is the most over-promised in the sector. A simple vibration threshold alert, reliably delivered to the right person's phone, prevents more failures in year one than any neural network. Earn the sophisticated models with two years of clean data; the cheap thresholds deliver the payback now. When you are ready to layer genuine machine learning on top, that is a deliberate step covered by a focused AI automation agency in London, not a box you tick on day one.
UK manufacturers can access meaningful grant funding for IIoT, with Made Smarter Adoption the most directly useful for SMEs because it can cover up to 50% of eligible digital technology project costs in supported English regions. This is real money that materially changes the business case, and it is under-claimed because owners assume grants are bureaucratic or only for large firms. They are neither, particularly Made Smarter, which was designed specifically for small and medium manufacturers adopting proven technology rather than inventing it.
The main funding routes to know in 2026:
The table below summarises who each route suits and the rough scale.
| Funding route | Best for | Typical support | Effort to apply |
|---|---|---|---|
| Made Smarter Adoption | SME adopting proven tech | Up to 50% matched grant + advice | Low to moderate |
| Made Smarter Innovation | Developing novel technology | Larger competitive grants | High |
| Innovate UK Smart Grants | Higher-risk R and D | Competitive project funding | High |
| HMRC R and D relief | Qualifying development spend | Tax relief on eligible costs | Moderate |
Our practical advice: start with your regional Made Smarter adoption team before you spend a penny. Eligibility and exact funding rates vary by region and change year to year, so check the current position for your area directly. The digital roadmap they provide free is genuinely useful even if you never claim the grant, because it forces you to define the business case properly. Be sceptical of any integrator who promises a grant will definitely cover your project; grant rates and budgets are finite and competitive, so treat funding as a welcome reduction, not a guarantee, and make sure the project stacks up on the numbers even at full cost.
One more honest point: write your business case so it survives without the grant. The strongest IIoT investments we have delivered would have paid back inside 18 months at full price. The grant simply accelerated the decision. If a project only works because someone else pays half, the underlying numbers are too weak and you should sharpen the scope before you start.
Industrial IoT can be secure, but security is the single most common reason UK manufacturers stall, with 67% citing it as the top barrier to expanding connected operations. That caution is justified. Connecting Operational Technology that was designed in an air-gapped era to networks and the cloud genuinely increases attack surface, and a compromised production system is far more dangerous than a compromised email account. It can halt output, damage equipment, or endanger staff. The answer is not to avoid IIoT; it is to do it under a proper security framework from the first sensor.
The UK regulatory and guidance landscape you must work within:
The practical defences that actually matter on a factory floor, in priority order:
The table below contrasts the risky shortcut with the secure approach for common decisions.
| Decision | Risky shortcut | Secure approach |
|---|---|---|
| Network design | Plug sensors into office LAN | Separate, segmented OT network |
| Device access | Keep default passwords | Unique credentials, least privilege |
| Cloud connection | Open inbound ports | Outbound-only, encrypted, edge-filtered |
| Vendor choice | Cheapest hardware | Vendor with patch commitment |
Our stance is uncompromising: if a supplier wants to put machine sensors on the same flat network as your office PCs, walk away. Segmentation is non-negotiable and inexpensive to do at the design stage, ruinous to retrofit after an incident. Security is not a feature you add at the end; it is the shape of the architecture you choose at the start.
Softomate delivers IIoT projects through a five-stage process that starts with one machine and a fixed-price discovery, never a giant rip-and-replace. We are a London-based automation and software agency in Stanmore, and our deliberate bias is to prove value cheaply on a single cell before you commit to a factory-wide programme. That protects your budget and means every later stage is justified by results you have already seen on your own shop floor.
The five stages:
We work to fixed quotes, not open-ended day rates, because manufacturers rightly hate uncertain bills. The timeline and indicative starting prices below cover a typical SME engagement.
| Stage | Typical timeline | Indicative starting price |
|---|---|---|
| Discovery and roadmap | 1 - 2 weeks | From £2,500 |
| Pilot build (one cell) | 3 - 6 weeks | From £12,000 |
| ERP/BI integration | 2 - 4 weeks | From £6,000 |
| Scale (per additional cell) | 1 - 2 weeks each | From £1,500 |
| Optimise and support | Ongoing | From £450/month |
A full first-phase engagement, discovery through to a pilot cell with live ERP integration, typically lands between £20,000 and £45,000 before any grant, with Made Smarter potentially halving the eligible portion in supported regions. Most clients then scale at a predictable per-cell cost because the hard architectural work is already done. Where the data needs to trigger actions across systems, raise tickets, alert supervisors, re-order stock, we build that on a proven automation backbone, and where bespoke dashboards or operator tools are needed we deliver them through our custom software development service in London. If you would value a straight conversation about whether your factory is ready, that starts on our contact page.
A sensible first phase covering one cell, from discovery through to live ERP integration, typically costs £20,000 to £45,000 before grants. Simple non-invasive sensing pilots can start nearer £12,000. Made Smarter Adoption can fund up to 50% of eligible costs in supported English regions, materially lowering your net spend.
Yes. Most UK factories connect legacy machines using non-invasive clamp-on sensors that read current, vibration and run state without touching the machine's controls, from around £400 to £1,200 per machine. You do not need to replace working equipment to start measuring OEE and downtime accurately.
The main routes are Made Smarter Adoption (matched grants up to 50% for SMEs adopting proven technology), Made Smarter Innovation and Innovate UK Smart Grants for novel R and D, plus HMRC R and D tax relief for qualifying development. Check your regional Made Smarter team for current eligibility.
Well-scoped pilots commonly pay back within 7 to 18 months. Predictive maintenance alone often covers the cost: cutting unplanned downtime by 60% on a single line worth £1,200 an hour in lost output can save nearly £29,000 a year against a pilot costing around £22,000.
Overall Equipment Effectiveness multiplies availability, performance and quality into one score. Most UK SMEs who measure it for the first time find a true OEE of 45 to 65%, far below the 85% they assumed. Closing that gap is where IIoT delivers most of its financial return.
It can be, with the right architecture. Around 67% of manufacturers cite security as their top barrier. The essential controls are network segmentation separating OT from office IT, removing default passwords, edge filtering and patching. The UK PSTI Act 2022 and NCSC operational technology guidance set the baseline you should follow.
Usually not. Most modern ERPs can ingest IIoT data through connectors or APIs, and integrating live machine output into your existing ERP is often higher value than the sensors themselves. If you are already planning an ERP refresh, design it to consume shop-floor data from day one rather than bolting it on later.
A focused pilot on a single cell, from sensors to a live dashboard, generally takes 3 to 6 weeks, with discovery adding 1 to 2 weeks beforehand and ERP integration another 2 to 4 weeks. You should see a meaningful downtime or OEE chart within the first fortnight of the build.
It means connecting your shop-floor machines (Operational Technology) to your business systems (Information Technology) so production data reaches the people making commercial decisions in seconds, not at the next weekly meeting. It is the core idea behind connecting your factory floor to business intelligence.
Start with OEE on your worst or most critical cell. It is cheaper to deploy, immediately exposes hidden losses, and builds the clean data foundation that genuine predictive maintenance later depends on. Treat machine-learning prediction as an earned upgrade once you have reliable condition data, not a day-one purchase.
IIoT connects your factory floor to business intelligence by sending sensor data through edge, network and platform layers into a dashboard and your ERP, turning gut feel into live numbers. For a UK SME, a one-cell pilot of £20,000 to £45,000, often halved by Made Smarter Adoption, typically pays back in 7 to 18 months through 30 to 50% less unplanned downtime and 10 to 20 points of OEE gain. Start small: retrofit your worst machine with non-invasive sensors from around £400, measure OEE, and let the first downtime chart justify the rest. Connect the data to your ERP because that is where the value compounds, and build security in from the first sensor under PSTI and NCSC guidance, never bolted on later. The factories pulling ahead in 2026 are not the ones with the most sensors; they are the ones that turned three numbers into daily decisions. The cheapest place to begin is one machine and one honest measurement.
Ready to see what your worst-performing machine is really costing you? Book a fixed-price discovery with our AI and automation agency in London and we will map a funded, low-risk path from your factory floor to a live business intelligence dashboard.
Written by Deen Dayal Yadav, Founder of Softomate Solutions, a London-based AI automation and software development agency in Stanmore (HA7). With over 12 years building software, automation and data systems for UK businesses, Deen helps manufacturers and SMEs connect operations to live business intelligence without ripping out what already works. Softomate Solutions is registered at Companies House and works with clients across London and the UK. Learn more about Softomate Solutions.
We protect the real names of all clients featured in examples and case studies. Every testimonial is from a real client.
More from the blog
Work with us
Book a free 30-minute discovery call with DD and get a personalised automation roadmap.
Softomate Solutions Limited
Deen Dayal Yadav
Online
We use essential cookies to keep the site running. With your permission, we also use analytics cookies to understand how visitors use our site so we can improve it. No data is sold. Privacy Policy
