ARIN6906 Chap.5 Sensing Environments and the Internet of Things
Sensing Environments and the Internet of Things
The Internet of Things turns objects and spaces into distributed systems of sensing, communication and action. The key loop runs from sensor to platform, model, decision and actuator, then back through environmental feedback. The analytical challenge is to compare useful anticipation with expanding visibility and administrative control.
What this chapter covers
- 01
Connected objects as network nodes
- 02
Sensors and environmental data
- 03
Anticipatory platforms and models
- 04
Actuators and automated intervention
- 05
Smart homes, logistics and cities
- 06
Data extraction, override and public governance
Worked example · free
Who controls a leak sensor in rented housing
- +1This checkpoint allocation is not an official University assessment scheme. Identify abnormal flow detection as the useful sensing function.
- +1Trace the chain from reading and threshold to valve action and staff repair priority.
- +1Explain how household routines become provider data and why false positives affect residents unevenly.
- +1Recommend notice, a rapid resident override, narrow secondary use and timely deletion.
Key terms
- Internet of Things
- Connected objects and environments that collect data, communicate and act through networked services.
- Sensor
- A component that translates a physical condition into data usable by a technical system.
- Actuator
- A component that changes a physical condition in response to a control signal.
- Anticipatory system
- A system that predicts conditions and intervenes before a directly observed event occurs.
- Override
- A practical route for an authorised person to interrupt or reverse automated action.
Sensing Environments and the Internet of Things FAQ
What changes when computation enters the environment?
Objects and spaces become able to sense conditions, communicate with platforms and trigger action. This can make maintenance or coordination faster, but it also makes occupants and resources more measurable and gives distant actors new capacity to intervene.
How should I analyse a smart-city claim?
Identify the public value being optimised, the data that represent it and the groups omitted by the metric. Then trace ownership and override. A city is not one machine, so competing needs and political contest cannot be reduced to dashboard efficiency.
Why is anticipation politically important?
An anticipatory system acts on a model of what is likely to happen. That model defines normal behaviour and can treat unusual needs as inefficiency or risk. Ask whose history trained the prediction and who bears the cost when it is wrong.
What is the difference between sensing and knowing?
A sensor produces a measurement under defined conditions. Turning that signal into knowledge requires calibration, interpretation and context. The gap matters because an environment may act on a proxy that is technically accurate yet socially incomplete, such as equating desk occupancy with productive work.
How can an IoT system redistribute agency?
Sensors make conditions legible, platforms compare them with thresholds and actuators intervene. Users may gain convenience while owners gain remote visibility and control. Mapping each step shows whether the system responds to occupants or primarily expands administrative capacity.
What should an override provide?
An override must be discoverable, timely and usable by the people affected. A nominal button that requires inaccessible credentials or responds after harm is not meaningful. Good governance also records why an intervention occurred and assigns someone to review repeated overrides as evidence of a flawed model.
Assessment move
Draw the control loop and annotate one actor at every stage. Add a second line showing who sees the data and a third showing who experiences the intervention. When a case promises efficiency, ask which public value has been compressed into the metric. End by judging whether the environment serves occupants or mainly makes them easier to administer.
Photograph or sketch one connected environment and mark every sensor, communication path, platform decision and actuator. Add the occupant who is affected but not represented by the available data. This missing actor often supplies the strongest critique. For presentation work, compare a reactive rule with a predictive one and explain how anticipation changes the burden of error.
For debate preparation, separate objections to data collection, automated intervention and institutional purpose; each requires a different safeguard. A precise recommendation should state what data are necessary, how long they remain useful, who can act on them and how an occupant can interrupt the loop. Create a sensor-to-action trace for each Internet of Things example.
Write down the physical condition being sensed, the signal actually measured, the inference produced, the party receiving it and the intervention that follows. Then add the people who did not install the device but become legible to it. This sequence reveals how an apparently objective reading can depend on placement, calibration, missing data and institutional interpretation.
Test the case under three failures: a false alarm, a missed event and a correct measurement used for an unrelated purpose. The comparison helps distinguish technical reliability from legitimate use. For assessment preparation, connect every proposed safeguard to a stage in the trace: calibration addresses measurement, access controls address circulation, appeal addresses classification and deletion limits future use.
End with a maintenance question. A sensing environment changes as buildings, residents and practices change, so responsibility includes recalibration, notice, contestability and eventual removal rather than only accurate installation.
