IoT

1. Meaning, components and architecture

The Internet of Things extends networked computing beyond conventional computers and smartphones to physical objects such as electricity meters, pumps, vehicles, factory machines and medical devices. These objects acquire information about themselves or their surroundings and communicate it for a useful purpose. Kevin Ashton is widely credited with coining the term in 1999 in the context of RFID-enabled supply chains. The defining feature is the integration of physical objects with digital sensing, communication and services, rather than simply the presence of an electronic component.

A typical architecture begins with a perception or device layer. Sensors measure variables such as temperature, pressure, motion, soil moisture or location. A network layer transports readings, sometimes through a gateway that aggregates data and translates between protocols. A processing layer stores and analyses information locally, at the network edge or in a cloud platform. The application layer presents results through dashboards, alerts or automated services. These layers describe functions; actual systems may combine them.

A sensor detects a physical condition, while an actuator changes it. For example, a moisture sensor measures soil conditions and an electrically operated valve controls irrigation. A monitoring-only system needs no actuator. Feedback-controlled systems compare observations with a target and trigger corrective action. This distinction matters because remote monitoring, remote control and fully automated operation represent different capabilities and different levels of risk.

• Gateway: an intermediary connecting devices or local networks to other networks and processing platforms.
• Embedded system: computing hardware and software built into a larger product; it becomes part of IoT when appropriately networked.

2. Connectivity and enabling technologies

IoT connectivity is selected according to range, power consumption, data rate, cost and reliability. Bluetooth Low Energy is useful for short-range, battery-powered wearables and beacons. Wi-Fi supports relatively high data rates within homes, offices and other local areas. IEEE 802.15.4-based technologies, including Zigbee, support low-power sensor networks and can use mesh arrangements. Wired links remain important in factories where predictable performance and resistance to radio interference may be valuable.

Low-power wide-area networks serve devices that transmit small amounts of data over comparatively long distances. LoRaWAN typically uses licence-exempt spectrum, subject to national rules, whereas NB-IoT is a cellular technology operating in licensed spectrum. LTE-M is another cellular option. Conventional mobile networks and satellite links also connect devices. No technology simultaneously maximises range, bandwidth, battery life and affordability: network design involves trade-offs and depends on local coverage.

IPv6 facilitates large-scale addressing but does not mean every sensor must have a publicly reachable Internet address. MQTT, commonly used in IoT, follows a publish–subscribe messaging model, often through a broker. CoAP is designed for constrained devices and networks. These are communication protocols, not sensors. Fifth-generation mobile networks can support massive machine-type communications and low-latency applications, but large numbers of IoT deployments work without 5G. Similarly, identification technologies such as RFID and NFC may contribute to connected services without independently providing continuous sensing.

3. Relationship with cloud, edge computing and artificial intelligence

IoT supplies observations from the physical world; cloud computing supplies remotely accessible computing and storage resources. A cloud platform can combine readings from thousands of devices and make them available across locations. However, transferring every reading to a distant data centre may consume bandwidth, expose sensitive information and introduce delay. Connectivity failures also make exclusive dependence on remote processing unsuitable for some applications.

Edge computing processes data on or near the device that generates it, such as inside an industrial controller or a local gateway. This can reduce latency and network traffic and allow essential functions to continue during an Internet outage. Fog computing usually describes a distributed arrangement of processing resources between end devices and a central cloud. Boundaries between these terms vary, but their shared principle is to avoid unnecessarily centralising every computing task.

Artificial intelligence can identify patterns in IoT data, predict machine failures or detect abnormal energy use. Nevertheless, a simple threshold-based temperature alarm is IoT without necessarily being AI. A digital twin is a digital representation of a physical entity or process that can be updated using operational data, including IoT measurements. It may support simulation and maintenance planning. Machine-to-machine communication overlaps with IoT but can also occur through closed, dedicated links without a broader Internet-based service.

4. Applications and Indian public-policy relevance

In agriculture, soil-moisture sensors, automated irrigation and livestock tracking can improve input management. Benefits depend on calibration, crop requirements, reliable connectivity and affordability; collecting data alone does not guarantee water savings. In healthcare, connected monitors can transmit physiological measurements and support follow-up outside hospitals. Such devices require appropriate clinical validation, security and safeguards against misleading readings. Consumer fitness wearables should not automatically be treated as diagnostic medical equipment.

Industrial IoT connects production equipment for condition monitoring, predictive maintenance and quality control. Urban applications include leakage detection, adaptive street lighting, waste-bin monitoring and air-quality measurement. Low-cost environmental sensors can increase spatial coverage, but calibration and quality assurance are necessary before treating their readings as equivalent to reference-grade measurements. Transport applications include fleet tracking and monitoring temperature-sensitive goods during distribution.

India’s Revamped Distribution Sector Scheme supports smart metering alongside distribution-system improvements. Connected meters can enable remote readings, consumption visibility and better identification of losses. Jal Jeevan Mission has also piloted sensor-based monitoring of rural water supply. Relevant institutions include the Department of Telecommunications, its Telecommunication Engineering Centre, and the Bureau of Indian Standards. TEC works on machine-to-machine and IoT technical requirements and standardisation, while interoperable standards can reduce dependence on a single supplier.

• Smart infrastructure should be assessed through service reliability, maintenance costs and measurable public outcomes, not merely the number of installed devices.
• Connectivity gaps, battery replacement and local technical capacity are important last-mile constraints.

5. Cybersecurity, privacy and sustainable deployment

IoT devices can remain operational for years despite limited memory, infrequent maintenance and weak update mechanisms. Default passwords, exposed network services and insecure software can allow attackers to compromise them. Large groups of compromised devices may form botnets that launch distributed denial-of-service attacks. In industrial and medical settings, malicious commands or corrupted readings may also affect physical safety, making IoT security more than a question of protecting stored information.

Core safeguards include unique credentials, strong authentication, encryption in transit, secure boot and authenticated software updates. Network segmentation limits movement from a compromised device to other systems. Organisations should maintain device inventories, define support periods and securely erase information before disposal. Essential systems also need fail-safe operation and manual override. Security must cover the entire chain: device hardware, firmware, gateways, applications, cloud accounts and third-party services.

Connected devices can reveal household routines, location and health conditions, sometimes through seemingly harmless readings. India’s Digital Personal Data Protection Act, 2023 provides a framework relevant to digital personal data processed through IoT, subject to applicable commencement provisions and rules. Purpose limitation, appropriate notices, data minimisation and retention controls are important governance considerations. Finally, interoperability problems, vendor lock-in, battery waste and electronic waste can weaken benefits. Procurement should therefore evaluate lifecycle security, repairability and environmentally sound disposal, not only initial purchase cost.

Previous year questions

No source-linked PYQ is available here yet. Use the practice questions below.

Practice questions