What can be expected from the future of IoT?

What can we expect from the future of IoT?
Hello, dear friends,
The real value of IoT's future lies in turning measurements into useful decisions, rather than simply connecting more devices. The Internet of Things is already used in many fields. Expectations include easier management, more local processing and better integration across manufacturers. However, wider adoption does not guarantee that every application will be economical or reliable.
Why does local processing matter alongside the cloud?
Edge computing processes data on a device or a gateway close to the equipment. The cloud can support long-term records, comparisons across sites and central management. Local processing can reduce connectivity requirements and transmitted data volumes. These complementary layers can allow some on-site functions to continue during an internet outage, provided the required logic, data and permissions genuinely reside locally.
For example, if each of 100 sensors generates 100 bytes of application data per second, the total is approximately 864 MB per day. This uses decimal megabytes and excludes protocol, encryption and retransmission overhead. Local summaries and event records may reduce transmission. However, which raw data to retain depends on the application: Excessive reduction may discard evidence needed to investigate a later fault.
How could artificial intelligence change IoT?
AIoT combines IoT data with artificial intelligence methods. Motor current, temperature and vibration trends may be analysed together to identify unusual conditions. However, detecting an anomaly is different from predicting when a failure will occur. Predictive maintenance requires suitable sensors, adequate sampling, operating context and validated failure history. The differences between IoT and IIoT become particularly important in industrial applications.
A new load pattern or relocated sensor may reduce a model's performance. False alarms, missed failures and changing operating conditions therefore need monitoring. AI does not automatically correct poor sensor installation or unreliable data. Where a simple threshold with a time delay meets the need, a more complex model may add unnecessary cost.
Will every IoT system need 5G?
No; connectivity should match the application. Choosing an internet connection for IoT involves coverage, data volume, latency, battery life and operating costs. Ethernet may suit a fixed machine, Wi-Fi or a low-power local network may suit a building, and suitable cellular connectivity may serve remote measurement points. 5G is one option, but its name alone provides no end-to-end latency guarantee. A video camera and a sensor sending a few readings each hour have different requirements.
What roles do MQTT and OPC UA play?
MQTT uses a publish–subscribe approach: Publishers send data to a message broker, which distributes it to relevant subscribers. OPC UA is an architecture that can describe industrial data structures, meaning and access methods. OPC UA PubSub can also use MQTT. The two technologies are not inherently mutually exclusive.
Device compatibility requires more than successful message delivery. Does “75” represent temperature or pressure? What is its unit, which machine produced it, and when was it measured? Data models, timestamps and quality information matter. MQTT delivery levels also do not, by themselves, verify that a physical command was executed correctly and safely. Communication delivery and process outcomes require separate consideration.
Will Matter solve every smart-home problem?
Matter is an IP-based standard intended to improve interoperability among compatible smart-home devices. Device types, supported features, software versions and controller platforms still need checking. Supporting the same standard does not mean every manufacturer-specific feature appears in every app. For everyday convenience, understandable setup, local control and behaviour during outages matter as much as feature counts.
Cybersecurity continues after the sale
IoT device security involves device identification, access permissions, data protection, trustworthy updates and awareness of security status. NIST IR 8259A describes baseline device capabilities in these areas. For maintenance and updates, check the support period, update-source verification and recovery arrangements after a failed update. A device working today does not establish that it can remain secure for years.
Assessing exposure to data breaches also involves network segmentation, limited privileges and controlled remote access. Given the cyber risks associated with continuous connectivity, industrial access should be restricted to necessary systems. Encryption helps, but does not by itself fix inappropriate permissions or outdated devices.
Privacy: Is more data always better?
Protecting personal privacy requires clarity about why data is collected, how long it is retained and with whom it is shared. Continuous monitoring concerns become particularly relevant to cameras, microphones and location tracking. Local processing can reduce transfers, but locally stored records still require protection. Avoiding unnecessary collection may be simpler than protecting additional data afterwards.
Industrial value: From an alarm to an appropriate response
If a pump's energy consumption rises, flow, pressure, operating time and process conditions should be considered together. Blockages, changing demand and incorrect settings may cause similar increases for different reasons. Remote monitoring is not, by itself, a diagnosis or a machinery safety function. Protection and control architectures must meet application requirements; an ordinary cloud dashboard should not be assigned a safety role. Existing PLCs, drives and sensors may be integrated through suitable gateways; complete replacement is not always necessary.
What could change in cities, agriculture, healthcare and homes?
Water-leak investigation, irrigation based on soil moisture and building energy savings can deliver measurable benefits. Sensor placement, calibration and maintenance responsibilities must nevertheless be defined. In healthcare IoT, data continuity, intended measurement use and professional assessment matter; not every consumer sensor is suitable for medical decisions.
For older adults and less experienced users, interfaces, controls and support should be accessible. Devices used by children also require attention to age-appropriate functions and limits on data collection. Technical sophistication does not automatically make a system easy for everyone to use.
How will costs, environmental impacts and skills change?
Long-term costs include networking, subscriptions, battery replacement, maintenance, data storage and migration, as well as the initial hardware price. Assessing environmental impact means considering achieved savings alongside electronics production, batteries and waste. Savings claims need measurement. Producing a chart does not itself save energy.
Employment effects cannot be reduced to universal claims that jobs will disappear or new ones will emerge. Tasks, training requirements and work organisation may change. Maintaining personal skills requires people to understand measurements, system limitations and appropriate fault responses. Successful automation also depends on human understanding.
How can you prepare today?
First identify the real need for IoT: Which loss should decrease, or which decision should improve? Establish a baseline in a small pilot, define a target and compare results under similar operating conditions. Assess data export, support duration, outage behaviour and migration to another system from the beginning. Preparation means building a sustainable system with measurable value, rather than buying every new connected device.
External sources and further reading
NIST SP 500-325 — Fog Computing Conceptual Model: How distributed processing relates to cloud computing.
OASIS — MQTT Version 5.0: The publish–subscribe messaging standard.
OPC Foundation — OPC UA: Industrial information models and communication architecture.
Connectivity Standards Alliance — Matter: The smart-home interoperability approach.
NIST IR 8259A — IoT Device Cybersecurity Capability Core Baseline: Baseline cybersecurity capabilities for IoT devices.
Future developments are expectations, not guaranteed outcomes or returns. You may share this article with attribution.