Internet of Things — Scientific Principles

Scientific Principles

An IoT architecture begins with a physical quantity and a sensor that measures it. The device communicates through a suitable network, often using an intermediary gateway. Processing may occur locally, at an edge system or in a remote service. An actuator changes a physical condition, such as operating a valve, when the design includes a control function. Sensing and actuation should not be confused.

The quality of a decision depends on the entire chain. A precise algorithm cannot correct an unrecognised sensor calibration error merely by receiving more readings. Missing data, delayed transmission and unsuitable placement can also change what the system appears to show. Validation should therefore include the measurement process, not just the dashboard.

Security requirements extend across a device's life. NISTIR 8259A, issued in May 2020, identifies six baseline capabilities: device identification, configuration, data protection, logical access to interfaces, software update and cybersecurity-state awareness. These capabilities support risk management but need adaptation to the device and use context. A guidance baseline is not proof of compliance with every Indian legal obligation.

Operation also requires ownership: who changes credentials, authorises an update, investigates an alert and replaces an unsupported device? Energy, connectivity and maintenance constraints can affect long-term performance. For a public project, the evaluation should connect the device data to a decision and an outcome rather than equating the number of installed sensors with service improvement.