How Will 6G Change IoT? 5G Promises vs Reality

We have been here before

The mobile industry introduces a new generation, produces diagrams filled with autonomous vehicles, surgeons, robots and holograms, and tells us that life is about to change. Investment follows. Trials begin. New phones acquire another symbol at the top of the screen. Then ordinary life carries on looking remarkably familiar.

5G was supposed to be the network for self-driving cars, remote surgery, intelligent factories, smart cities and billions of connected devices. By 2026, its most visible achievement for many UK users is a faster phone connection in some locations and a 5G icon whose practical meaning varies considerably. That does not mean 5G failed. It means the public story compressed a complicated, decade-long transformation into a product launch.

Now 6G is entering the same dangerous stage, with talk of intelligent networks, radio that can sense, programmable surfaces, centimetre-level positioning, batteryless sensors and distributed artificial intelligence. So before we ask how 6G will change IoT, the useful questions are these: who is 6G actually for, what happened to the industrial and automotive promises made for 5G, and why are countries still building 450 MHz and 600 MHz networks while engineers discuss frequencies above 100 GHz?

The answers begin with an awkward truth. Wireless networks do not advance in a straight line from slow to fast. A remote electricity meter and an immersive headset do not need the same network. Neither does a driverless vehicle, a factory robot, a soil sensor or a mobile phone. The future of wireless is not one network doing everything. It is a collection of networks becoming better at working together.

What was 5G originally supposed to be?

The formal vision for 5G was considerably more sensible than much of the advertising built around it. Under the ITU IMT-2020 framework, 5G had three broad families of use.

Enhanced Mobile Broadband (eMBB)

The obvious consumer proposition: faster downloads, higher capacity and better performance in busy places. It supported mobile video, cloud applications and fixed wireless broadband, while creating room for augmented and virtual reality.

Ultra-Reliable and Low-Latency Communications (URLLC)

Intended for applications where a delayed or missing packet could have serious consequences, such as industrial control, robotics, remote operation and parts of connected transport. This was not simply faster internet. It required predictable performance, very low delay and high availability across the whole service path.

Massive Machine-Type Communications (mMTC)

The IoT proposition: very large numbers of connected sensors and machines, many sending small amounts of data and needing long battery life. NB-IoT and LTE-M became important here, although both grew out of the LTE family rather than requiring a conventional high-performance 5G modem.

That three-part vision already revealed the difficulty. The industry used one generation number to describe three radically different jobs. A phone user wanted hundreds of megabits. A robot wanted predictable milliseconds. A water meter wanted to send a few bytes from a basement for ten years without anyone changing its battery. There was never going to be one radio configuration ideal for all three.

What were we told 5G would do?

Around the end of the 2010s, the careful distinctions were largely lost. 5G became attached to almost every fashionable technology story: fleets of fully autonomous cars, surgeons operating remotely across continents, smart cities reacting instantly to their citizens, factories filled with untethered robots, drones beyond visual line of sight, holographic meetings, millions of sensors per square kilometre and household broadband delivered without fixed lines.

UK government programmes explored connected vehicles, industrial automation, ports, logistics, broadcasting, agriculture and healthcare. Testbeds demonstrated valuable technical capabilities. They also encouraged a tendency to describe anything connected wirelessly as a 5G use case, even when Wi-Fi, fibre or 4G could already do much of the job. The phrase "enabled by 5G" became wonderfully elastic. Sometimes 5G was essential, sometimes it improved an application, and sometimes it was simply the network available to a funded trial.

Did 5G fail to deliver?

No, but it delivered in a less theatrical form than the publicity suggested. 5G has increased mobile capacity, improved download performance, and supported fixed wireless access, broadcasting, venue connectivity, private network trials and industrial projects. Operators modernised their radio networks and gained experience that will carry into later generations. Those are real achievements.

The disappointment came from confusing three different stages: a capability appearing in a technical standard, an operator implementing it in a live network, and a customer finding a reliable and economical reason to buy it. The gap between those stages can be measured in years.

Britain's first public 5G services were predominantly 5G Non-Standalone (5G NSA). They used a 5G radio connection while still depending on parts of the existing 4G core. This was a practical way to introduce higher speeds, but it did not immediately provide the complete low-latency, slicing and service-control proposition associated with a fully standalone system. Understanding the difference between 5G NSA and 5G Standalone in the UK is the first step to reading these numbers correctly.

Ofcom's Connected Nations 2025 report, published in November 2025 and based on July 2025 data, found 5G coverage outside premises from at least one operator across 94% to 97% of the UK, and 5G Standalone reaching 83% at the high-confidence level. Yet 4G still carried 72% of total monthly mobile data traffic. 5G Standalone accounted for roughly 31% of 5G traffic and only about 9% of all mobile data.

That is not a contradiction. It is what a generational transition actually looks like. Coverage, compatible devices, tariffs, traffic and business adoption all move at different speeds. The nation had moved a long way beyond the 2019 launches, but it had not suddenly become the uniformly sliced, deterministic industrial platform suggested by the early brochures.

What happened to the self-driving cars?

The self-driving-car story is the clearest example of the difference between a useful network and a necessary dependency. A safe autonomous vehicle cannot rely on uninterrupted mobile coverage to see a pedestrian, recognise a red light or decide whether to brake. The safety-critical driving system must make essential decisions using sensors and computing aboard the vehicle.

Cellular connectivity can still make the vehicle better. It can provide high-definition map updates, traffic and hazard information, fleet management and diagnostics, software updates, cooperative perception, remote assistance in unusual circumstances and communication with road infrastructure. Those are substantial roles. They are not the same as 5G driving the car.

The UK has developed test facilities, trials and legislation for automated vehicles, and small-scale commercial passenger pilots are now emerging, while the wider framework under the Automated Vehicles Act is still being implemented. The delay was never caused simply by inadequate mobile coverage. Autonomous driving also depends on safety validation, sensors, machine perception, insurance, liability, mapping, regulation, cost and public acceptance. 5G was one part of a much larger machine. Marketing sometimes presented it as the ignition key.

The lesson from private 5G

Industrial 5G was another credible idea that met the less glamorous realities of procurement. A factory may value private 5G for controlled coverage, mobility, security, predictable quality of service and the ability to connect machines across a large site. Ports, mines, warehouses and large campuses can present equally strong cases.

But private 5G competes with technologies that already exist. Industrial Ethernet is dependable. Wi-Fi is familiar and inexpensive. Existing 4G equipment is widely available. LoRaWAN can connect low-power sensors over a large site at modest cost. A private mobile network adds radios, spectrum, a core, SIM provisioning, compatible devices, installation and specialist management. The business case must therefore solve an operational problem worth paying for.

The market is real but measured. In April 2026, Vodafone launched what it described as the UK's first commercial 5G network slicing service, aimed initially at guaranteed performance across defined business sites such as stadiums, campuses and logistics hubs. That is genuine progress toward the slicing proposition promised years earlier, and it is arriving as a bookable product rather than a brochure claim, which is exactly the distinction that matters.

What is 6G?

6G is the next generation of international mobile telecommunications, formally described by the ITU as IMT-2030. It is not yet a finished standard. The ITU is defining the requirements against which candidate radio technologies will be evaluated, while 3GPP develops the specifications. Release 21 is expected to be the first normative 3GPP 6G release, with initial commercial deployments anticipated around 2030.

The clearest useful definition is that 6G is intended to combine communications, sensing, positioning, computing and artificial intelligence within a common wireless platform. The ITU has organised its 6G vision around six usage scenarios:

• Immersive Communication
• Hyper-Reliable and Low-Latency Communication
• Massive Communication
• Ubiquitous Connectivity
• Artificial Intelligence and Communication
• Integrated Sensing and Communication

The first three extend the broadband, low-latency and massive-IoT ambitions of 5G. The final three make the more fundamental change visible. The 6G network is not only expected to connect devices. It may help locate them, perceive their surroundings, allocate computing tasks and even modify the radio environment through which they communicate.

Who is 6G actually for?

The first beneficiary will not necessarily be a person buying a new phone.

Mobile network operators

Operators need greater spectral efficiency, lower energy use, more automation and simpler operation. A network that can predict faults, optimise radio resources and place workloads intelligently may reduce costs as well as add services.

Industry and infrastructure

Factories, utilities, transport systems, ports, logistics and public-safety organisations may benefit from communications combined with positioning, sensing and edge computing.

Machines and autonomous systems

Robots, vehicles, drones and industrial machines need to understand their surroundings, exchange selected information and coordinate actions. 6G is being designed with machine-to-machine collaboration more explicitly in mind.

AI applications

An AI workload may be divided between a small endpoint, a nearby edge computer and a larger cloud platform. The network could help decide where inference should happen, which data needs transmitting and how quickly a result must return.

Consumers

Consumers will receive faster and more consistent mobile broadband. Immersive entertainment and new interfaces may eventually matter, but they are unlikely to be the only, or even the earliest, reason for building 6G.

How will 6G change IoT?

The biggest changes will not come from placing a 6G modem in every temperature sensor. They will come from expanding what the surrounding network can do.

1. The network becomes a sensor

Integrated Sensing and Communication (ISAC) lets radio infrastructure communicate and sense at the same time. By analysing signals and their reflections, a network could estimate range, direction and velocity, detect objects or movement, improve positioning, or recognise that physical conditions have changed. That creates the possibility of device-free sensing, where a person, vehicle or object is detectable without carrying a tracker. For IoT, some sensing moves away from individual battery-powered endpoints and into shared infrastructure.

2. The physical environment becomes programmable

Reconfigurable Intelligent Surfaces use arrays of controllable elements to alter how radio waves are reflected or transmitted. A wall could stop being merely an obstruction and become part of the coverage system, redirecting energy toward a difficult area or supporting sensing as people and machinery move. The grand idea is a smart radio environment, where parts of the building join the network rather than fighting the signal.

3. Positioning becomes part of the service

Early ITU research targets envisage positioning accuracy measured in centimetres for applicable scenarios. That will not be available everywhere, and not every sensor needs it. Where available, integrated high-precision positioning could support robotics, warehouses, ports, vehicles, asset tracking and digital twins without a separate location system for every application.

4. Communication becomes goal-oriented

Conventional networks transport bits without caring what they mean. Semantic and goal-oriented communication asks a different question: what does the receiving application actually need? A camera might upload continuous video, or an intelligent edge system might transmit only that a vehicle entered a restricted area, its direction, an identifier and a short clip. That reduces traffic, storage, latency and energy. It also adds a risk: if the local system misreads the event, the discarded information is no longer there for a human to inspect.

5. Computing becomes part of connectivity

6G increasingly treats communication and computation as related resources. A device may collect data, do a little local processing, send part of a task to an edge node and use the cloud only when necessary, with the network allocating those resources by latency, energy, privacy and cost. For constrained hardware, that could mean more useful intelligence without turning every endpoint into an expensive computer.

6. Terrestrial and satellite networks converge

Non-Terrestrial Networks are already part of 5G evolution, covering satellites and high-altitude platforms. Under 6G, the ambition is closer integration, so a remote asset uses terrestrial service where available and satellite elsewhere, with less distinction visible to the application. For agriculture, maritime, energy, environmental monitoring and logistics, that matters more than another jump in urban download speed. The UK has already seen the first steps toward this with operator satellite and NTN activity for M2M.

7. Some devices may operate without conventional batteries

Ambient IoT and backscatter communication aim to support very simple devices powered by harvested energy from light, heat, vibration or radio waves. Such devices would send tiny amounts of data, and their value would come from low cost, minimal maintenance and the ability to attach intelligence to objects that cannot justify a cellular modem and battery. The phrase "zero-energy device" needs care: energy still comes from somewhere. It usually means avoiding a replaceable battery, not breaking the laws of physics.

Why are we still developing 450 MHz and 600 MHz networks?

Because frequency is not a league table in which the highest number wins. 450 MHz and 600 MHz describe frequency bands, not generations. LTE can operate at low frequencies. So can 5G, and future 6G systems are expected to use a wide range of spectrum.

Lower-frequency radio generally travels further, bends around obstacles more effectively and penetrates buildings better than very-high-frequency radio. Higher frequencies can offer much wider channels and enormous capacity, but usually over shorter distances and with greater sensitivity to walls and foliage. The free-space wavelength shows the physical difference.

Why 450 MHz matters

LTE450 is particularly attractive to utilities, public-safety organisations and operators of critical national infrastructure. Electricity grids, water networks and pipelines cover large areas, and their endpoints may sit in basements, metal cabinets or remote substations. They usually need dependable coverage and modest data rather than consumer broadband speeds. A 450 MHz network can cover a large area with fewer sites, and dedicated spectrum can provide greater control, resilience and longevity. For these users, ten megabits delivered reliably in the right place may be worth far more than a gigabit beside a city-centre lamppost.

Why 600 MHz matters

The same logic favours 600 MHz as a coverage layer. A low-band 5G signal may not produce the most spectacular speed test, but it can make the service available across rural areas and inside buildings, while higher bands add capacity where demand is dense. The principle survives into 6G. The ITU envisages IMT-2030 using frequencies from below 1 GHz to bands above 100 GHz. Low frequencies provide the coverage foundation, mid-bands provide the everyday balance, and very high frequencies provide specialised capacity and sensing. 6G does not abolish radio physics.

Does a low-frequency network mean a low-bandwidth network?

Not automatically. Frequency is the position of the channel in the spectrum. Bandwidth is the width of spectrum allocated to the signal. A 600 MHz 5G carrier can use modern coding, MIMO and network architecture, yet its channel may still be narrower than a carrier at 3.5 GHz because less contiguous spectrum is available down low.

The distinction matters. A lower carrier frequency usually improves reach and penetration. A wider channel allows higher potential throughput. Modulation, signal quality, antenna configuration and network loading also affect performance. Calling LTE450 a slow old network therefore misses its purpose. Its limited capacity can be a rational trade-off for coverage, resilience, ownership and long asset life.

Will 6G replace 5G, 4G, Wi-Fi and LoRaWAN?

No, not quickly, and probably never in the absolute sense. 4G will remain in service for years because millions of phones, routers, vehicles, meters, alarms and industrial systems depend on it. 5G will still be expanding when the first 6G networks appear. Wi-Fi will keep dominating local connectivity. LoRaWAN and other low-power systems will remain suitable where tiny messages, private gateways and long battery life matter more than mobility or throughput.

6G will enter an already crowded environment containing 4G LTE, LTE-M and NB-IoT, 5G NSA and 5G Standalone, RedCap and eRedCap, private mobile networks, Wi-Fi, LoRaWAN and other LPWAN, satellite IoT, and dedicated systems such as LTE450. The job of an IoT designer will be to choose intelligently between them, not to assume the newest generation is automatically correct.

What should 6G learn from 5G?

Stop presenting demonstrations as deployments. A laboratory system, a controlled testbed and a nationwide commercial service are different things. Coverage, device availability, economics, regulation and operational support matter as much as a successful radio demonstration.

Explain which part of the system requires 6G. If an application also needs fibre, edge computing, AI, specialist sensors and new regulation, it should not be presented as something 6G produces single-handedly.

Publish meaningful service characteristics. An industrial customer needs availability, latency under load, uplink capacity, geographic coverage, recovery time, device lifecycle and contractual support. A theoretical peak speed is rarely the deciding factor.

Design for migration. A deployment installed in 2030 may still be operating in 2045. Hardware, SIMs, eSIM profiles, antennas and management platforms should be designed around network change rather than one permanent operator and radio generation.

Treat sensing privacy as a first-order problem. A network that can detect movement, posture, location and unconnected objects creates real benefits and real surveillance capability. That cannot be dismissed as a minor software-policy issue.

Be honest about energy. Better efficiency per transmitted bit does not automatically mean lower total consumption. More sites, more computing and vastly more traffic can cancel the gain. Sustainability must be measured across infrastructure, devices and their whole lifecycles.

So how will 6G really change IoT?

6G could make the wireless network aware of far more than connected devices. It could sense physical surroundings, provide precise positioning, coordinate computing, combine terrestrial and satellite coverage, support energy-harvesting endpoints and use programmable surfaces to reshape difficult radio environments. That is genuinely different.

But most IoT deployments will not need every 6G capability. A tank-level sensor may still want a few bytes, excellent penetration and ten years of battery life. A utility may still prefer a privately controlled 450 MHz network. A factory may choose Wi-Fi, private 5G or wired Ethernet. A vehicle must still stay safe when its connection disappears. The future is not a universal march from 4G to 5G to 6G with every older technology swept aside. It is a layered wireless environment in which different frequencies, technologies and ownership models serve different operational needs.

5G taught us that a new generation can be technically successful while its most dramatic predictions stay years away. The lesson for 6G is not to abandon ambition. It is to describe that ambition honestly. That would not merely be faster IoT. It would be the beginning of an intelligent wireless environment.

Frequently asked questions about 6G and IoT