Healthcare and Smart-City Applications

Introduction

The Internet of Things (IoT) becomes economically and socially significant when connected devices are embedded in environments where decisions, services and resources must be coordinated continuously. Healthcare and smart-city systems are two important application domains because both involve complex infrastructures, multiple stakeholders, sensitive data and consequences for human wellbeing.

In healthcare, IoT connects medical devices, patients, clinicians, hospitals, laboratories, pharmacies and insurers. Connected devices can collect physiological information, support diagnosis, automate clinical operations and enable care outside hospitals. In cities, IoT connects transport systems, utilities, buildings, public infrastructure, emergency services and environmental sensors. The objective is not simply to install sensors, but to convert data into safer, more efficient, inclusive and sustainable services.

These applications also expose the limitations of technology-centred thinking. A technically reliable sensor may still produce poor outcomes if patients cannot use it, clinicians cannot interpret its data, agencies cannot share information, or citizens do not trust the system. Consequently, managers must evaluate IoT as a socio-technical system involving technology, people, processes, institutions, regulation and finance.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Explain the architecture and business value of connected medical devices and remote patient monitoring.
2. Analyse applications of IoT in clinical and healthcare operations.
3. Evaluate data quality, interoperability, cybersecurity, privacy and safety risks in connected healthcare.
4. Explain smart-city applications in mobility, utilities, infrastructure, public safety and environmental management.
5. Assess business models, governance arrangements and performance measures for healthcare and smart-city IoT programmes.
6. Recommend implementation approaches that balance innovation, inclusion, resilience and public value.

1. Connected Healthcare as a Socio-Technical System

Connected healthcare IoT consists of sensing, communication, data management, analytics and action. A wearable blood-pressure monitor, for example, measures a patient’s condition; transmits readings through a phone or gateway; stores the data in a clinical platform; applies rules or analytics; and presents an alert or trend to a clinician. The final value arises only when an authorised person takes an appropriate action.

A useful healthcare architecture contains five layers:

This layered model clarifies accountability. A connectivity failure is different from a faulty sensor, an incorrect algorithm or an ignored alert. Healthcare leaders should therefore evaluate the entire chain rather than purchase devices in isolation.

Connected medical devices

Connected medical devices are instruments that measure, transmit, receive or act on health-related information. They include bedside monitors, smart inhalers, implantable cardiac devices, connected imaging systems, wearable activity monitors and smart infusion pumps. Some devices are regulated as medical devices because their intended use affects diagnosis, prevention, monitoring or treatment. Others are wellness devices and may not satisfy the same evidential or regulatory requirements.

The distinction matters. A consumer fitness band may estimate heart rate for personal awareness, whereas a clinically validated cardiac monitor may generate information used in treatment decisions. Managers must examine intended use, validation evidence, calibration, maintenance, software updates, data ownership and integration with clinical workflows.

The principal value propositions are:

• earlier detection of deterioration;
• improved adherence to treatment;
• reduced manual observation and transcription;
• more personalised care;
• improved asset utilisation;
• lower avoidable admissions and readmissions; and
• richer evidence for service planning and population health.

These benefits are conditional. More data does not automatically produce better care. Excessive alerts can create alarm fatigue, while missing data can create false reassurance. A device programme should therefore specify the clinical decision it supports, the responsible professional, the response time and the escalation procedure.

Remote patient monitoring

Remote patient monitoring (RPM) is the structured collection and clinical use of patient data while the patient is outside a conventional care facility. Typical targets include blood pressure, blood glucose, oxygen saturation, body weight, temperature, heart rate and symptoms. RPM is especially relevant to chronic diseases such as diabetes, hypertension, chronic obstructive pulmonary disease and heart failure.

A basic RPM pathway is:

1. The patient uses a sensor or measurement device.
2. Data are transferred to a gateway or application.
3. The platform checks completeness, plausibility and risk thresholds.
4. Clinicians review dashboards, trends or alerts.
5. The care team contacts the patient or changes the care plan when indicated.
6. Outcomes and exceptions are documented for evaluation.

RPM should not be confused with teleconsultation. Teleconsultation primarily provides communication between patient and professional, whereas RPM provides continuous or repeated measurement that can inform communication and intervention. They are often combined.

A simple operational measure is adherence:

The measure assumes that the expected measurement schedule is clearly defined and that a received measurement is technically valid. For example, if a programme expects 30 daily readings and receives 27 valid readings, adherence is 90%. Adherence alone does not establish clinical effectiveness; a patient may submit all readings without improving health outcomes.

Other relevant indicators include alert precision, response time, hospital utilisation, patient-reported outcomes, equity of access, device failure rate and total cost per enrolled patient. Clinical outcomes should be compared with an appropriate baseline or control group where feasible.

Data quality and clinical safety

Healthcare IoT data must be assessed for accuracy, completeness, timeliness, consistency and provenance. A reading without a timestamp, patient identity or device context may be unusable. Data quality problems can result from sensor drift, poor placement, low battery power, connectivity gaps, duplicate records or incorrect patient-device association.

A practical quality-control process includes device validation, patient training, automated plausibility checks, exception handling, calibration schedules and human review. Algorithms should communicate uncertainty rather than present estimates as facts. Where artificial intelligence is used, organisations should examine training-data bias, performance across demographic groups, explainability, model drift and the consequences of false positives and false negatives.

Clinical safety requires risk management throughout the device lifecycle. Important controls include:

• authentication and least-privilege access;
• encryption during transmission and storage;
• secure software development and patch management;
• audit logs and incident response;
• fail-safe operating modes;
• documented downtime procedures;
• human oversight of high-risk decisions; and
• post-market monitoring of adverse events.

Privacy is especially important because health data are sensitive and may reveal behaviour, location, disability, reproductive information or genetic risk. Consent should be meaningful, specific and understandable. Organisations should collect only necessary data, limit retention, explain secondary uses and provide mechanisms for access and correction. Legal obligations vary by jurisdiction, but sound governance also requires ethical justification beyond minimum compliance.

2. IoT in Healthcare Operations

IoT extends beyond direct patient monitoring. Healthcare organisations are asset-intensive operations involving beds, operating theatres, pharmaceuticals, laboratories, vehicles, energy systems and staff workflows.

Asset and equipment management

Radio-frequency identification, Bluetooth beacons and networked equipment can locate devices, monitor utilisation and support preventive maintenance. A hospital may discover that infusion pumps are unevenly distributed: one ward experiences shortages while another holds idle inventory. Location and utilisation data can improve allocation, reduce search time and support replacement planning.

Predictive maintenance uses equipment condition data to identify likely failures before they disrupt care. However, maintenance decisions must distinguish between a statistical warning and a confirmed fault. A critical device should not remain in service merely because its failure probability is estimated to be low.

Medication and cold-chain monitoring

Connected temperature sensors can monitor vaccines, blood products and medicines requiring controlled storage. Alerts can be configured for temperature excursions, door openings or power failures. The business case includes reduced waste, improved traceability and stronger compliance evidence. Yet sensors require calibration, backup power and escalation protocols; otherwise an alert may be generated without a practical response.

Facilities and workforce operations

Occupancy sensors, environmental monitors and connected building systems can manage heating, ventilation, lighting and space allocation. In hospitals, indoor air quality and pressure relationships may be important for infection control. Staff-location systems may assist emergency response, but they also raise labour privacy and surveillance concerns. Deployment should therefore involve workforce consultation, purpose limitation and transparent policies.

3. Smart-City Applications

A smart city uses digital technologies, data and institutional coordination to improve urban services and public value. IoT is an enabling infrastructure, not a complete definition of a smart city. A city can have millions of sensors yet remain poorly governed, exclusionary or inefficient.

A smart-city system normally combines field devices, communications networks, shared data platforms, analytics, operational control centres and public interfaces. Interoperability is essential because transport, water, energy, emergency and environmental systems are commonly operated by different agencies and suppliers.

Mobility and transport

Mobility IoT includes traffic counters, connected traffic signals, parking sensors, fleet telematics, public-transport trackers, bicycle-sharing systems, electric-vehicle charging stations and road-weather sensors. Applications include:

• adaptive signal timing based on traffic conditions;
• real-time bus and rail information;
• predictive maintenance of vehicles and tracks;
• dynamic parking guidance;
• congestion and emissions management;
• logistics-route optimisation; and
• priority routing for emergency vehicles.

A city should evaluate mobility systems against accessibility, travel time, reliability, safety, emissions and equity—not merely sensor counts. Dynamic pricing can reduce congestion, but it may disadvantage lower-income travellers unless revenue and exemptions are designed fairly.

Utilities and energy

Smart meters and network sensors provide more frequent information about electricity, gas and water consumption. Utilities can detect leaks, identify outages, forecast demand and support demand response. In a demand-response programme, customers may reduce or shift consumption during stressed periods in response to information or price signals.

For water systems, pressure and flow sensors can identify abnormal patterns associated with leakage or pipe failure. For electricity systems, distributed energy resources such as solar panels, batteries and electric vehicles require monitoring and coordinated control. Cybersecurity is critical because an attack on utility control systems can affect public health and economic continuity.

Infrastructure and buildings

Structural-health monitoring uses sensors to observe vibration, strain, corrosion, temperature or movement in bridges, tunnels, dams and buildings. Data can support inspection prioritisation and maintenance planning. A sensor does not replace engineering inspection: it provides evidence that must be interpreted in relation to design limits, environmental conditions and the asset’s history.

Smart buildings use connected heating, ventilation, lighting, access control and occupancy systems. Energy savings can be substantial, but poorly designed automation may reduce comfort, accessibility or resilience. Building managers should specify manual overrides and maintain safe operation during network or power failures.

Public safety and emergency management

IoT supports emergency call systems, flood gauges, wildfire detection, connected street lighting, gunshot-location systems, public warning displays and responder wearables. During a flood, river-level sensors can provide earlier warning, while traffic data can identify evacuation routes and congestion.

Public-safety applications require strict governance. Facial recognition, pervasive location tracking and predictive-policing tools may create discrimination, chilling effects and due-process concerns. A city should establish legal authority, necessity, proportionality, retention limits, independent oversight and appeal mechanisms before deployment.

Environmental management

Air-quality, noise, heat, water-quality and waste sensors provide fine-grained evidence about urban conditions. Smart waste systems may use fill-level sensors to adjust collection routes. Urban heat sensors can identify neighbourhoods requiring shade, cooling centres or tree planting. Environmental data can also support climate adaptation and emissions reporting.

Measurements must be interpreted carefully. Sensor placement affects representativeness; low-cost sensors may require calibration; and data gaps may occur in communities with weaker connectivity. Environmental monitoring should therefore combine sensor data with validated reference stations and community knowledge.

4. Business Models, Governance and Evaluation

Healthcare and city IoT programmes involve complex value networks. Suppliers may sell devices, connectivity, platforms, analytics or managed services. Public agencies may procure systems, while hospitals, utilities, citizens and insurers realise different benefits. Contract design should address service levels, cybersecurity obligations, data portability, ownership, intellectual property, audit rights, exit arrangements and liability.

Useful evaluation categories include:

A business case should include total cost of ownership: devices, installation, connectivity, integration, training, maintenance, cybersecurity, data governance and eventual replacement. Benefits should be risk-adjusted and should distinguish financial savings from broader public value.

Interoperability is a strategic issue. Open standards and well-defined application programming interfaces can reduce vendor lock-in and permit data exchange. However, interoperability without access control can increase exposure. Data architecture should therefore combine standardisation with identity management, consent controls, segmentation and monitoring.

5. Implementation Framework

A responsible implementation can follow seven stages:

1. Define the problem. Specify the service failure or clinical decision to improve.
2. Map stakeholders and harms. Include patients, citizens, clinicians, operators, regulators, suppliers and vulnerable groups.
3. Design the minimum viable data system. Collect only information needed for the intended decision.
4. Pilot in realistic conditions. Test connectivity, usability, workflow, safety and equity—not only technical performance.
5. Evaluate outcomes. Compare results with baseline conditions and document unintended effects.
6. Scale with governance. Establish standards, procurement controls, cybersecurity, training and accountability.
7. Review and retire. Monitor drift, update systems and decommission devices securely.

This approach prevents the common error of scaling a technically impressive pilot that has no sustainable operating model. It also recognises that adoption is an organisational change problem. Clinicians may reject systems that increase documentation burden; residents may reject surveillance that lacks legitimacy; and operators may bypass unreliable alerts.

Illustrative Case: Connected Heart-Failure Management in a Regional Health Network

Background

A regional health network experienced frequent readmissions among patients with heart failure. Patients often deteriorated at home because weight gain, breathlessness and medication problems were identified late.

Problem

The network considered issuing connected weighing scales, blood-pressure monitors and symptom questionnaires. The initial proposal focused on purchasing devices, but analysis showed that the central challenge was response capacity: no team had been assigned to review alerts consistently.

Approach and solution

The network redesigned the programme around a care pathway. Patients received devices and training before discharge. Measurements were transmitted through a mobile gateway. A rules engine identified clinically meaningful changes, but alerts were tiered to reduce alarm fatigue. A nurse team reviewed high-priority cases during defined hours, contacted patients, and escalated cases to physicians. Patients without reliable broadband received cellular gateways or telephone-supported alternatives.

The programme established device replacement procedures, data-quality checks, documented consent, access controls and a downtime process. Evaluation included readmissions, emergency visits, adherence, nurse response time, patient experience and equity by age and socioeconomic status.

Outcome

The network reported improved visibility of patient trends and faster contact with some high-risk patients. It also found that adherence was lower among patients with limited digital confidence and that false alerts increased when patients used scales incorrectly. Additional coaching and simpler interfaces were introduced. The case demonstrates that clinical workflow, inclusion and measurement quality determine value more strongly than device quantity.

Learning points

1. A connected device must be linked to a defined decision and accountable response.
2. Alert design should reflect clinical capacity and risk tolerance.
3. Digital inclusion is a performance requirement, not an optional social programme.
4. Evaluation must include outcomes, workload, equity and unintended consequences.

Challenges and Contemporary Developments

Major challenges include fragmented legacy systems, uncertain return on investment, cybersecurity threats, unreliable connectivity, vendor dependence, skills shortages and regulatory complexity. Rural and low-income communities may receive fewer benefits because infrastructure and digital access are uneven.

Artificial intelligence is increasingly used for anomaly detection, demand forecasting, triage support and predictive maintenance. Its value depends on representative data, continuous validation and human accountability. Generative AI may help summarise sensor data or draft operational reports, but it must not obscure source evidence or make unsupervised high-risk decisions.

Digital twins—virtual representations linked to physical assets—are emerging in hospitals, utilities and city infrastructure. They can support scenario analysis, such as testing traffic changes or energy strategies. Their reliability depends on model assumptions and data freshness; a digital twin is not automatically a complete or accurate replica.

The strategic direction is toward interoperable, resilient and human-centred systems. The most mature programmes treat data as shared infrastructure, establish clear public-interest purposes and design for failure, accessibility and accountability from the beginning.

Chapter Summary

Healthcare IoT connects medical devices, patients, clinicians and operational assets so that data can support earlier intervention, personalised care and more efficient services. Remote patient monitoring is valuable only when measurement, transmission, interpretation and clinical response form a reliable pathway. Data quality, cybersecurity, privacy, interoperability, clinical safety and digital inclusion are central management concerns.

Smart-city IoT applies connected sensing and control to mobility, utilities, infrastructure, public safety and environmental management. Its success should be evaluated through outcomes such as safety, reliability, sustainability, accessibility, equity and resilience rather than through the number of deployed sensors. Effective programmes begin with clearly defined problems, use proportionate data, involve stakeholders, pilot realistically and establish lifecycle governance.

Key Terms

Connected medical device: A device that collects, communicates or acts on health-related information.

Remote patient monitoring: The structured collection and clinical use of patient data outside a conventional care facility.

RPM adherence: The proportion of expected, valid patient measurements actually received.

Alert fatigue: Reduced attention or responsiveness caused by excessive or low-value alerts.

Interoperability: The ability of systems to exchange and use information effectively.

Clinical workflow: The sequence of tasks and decisions through which care is delivered.

Sensor calibration: Adjustment or verification of a sensor against a known reference.

Digital inclusion: Equitable ability to access and benefit from digital services.

Smart city: An urban system using data, technology and governance to improve public services and outcomes.

Smart mobility: Connected management of transport, travel demand, vehicles and infrastructure.

Smart meter: A metering device that records and communicates consumption data.

Demand response: Adjustment of consumption in response to system conditions or incentives.

Structural-health monitoring: Sensor-based observation of an asset’s condition and behaviour.

Predictive maintenance: Maintenance planned using evidence about likely equipment deterioration or failure.

Environmental sensing: Measurement of conditions such as air quality, noise, heat or water quality.

Digital twin: A data-linked virtual representation of a physical asset or system.

Data provenance: Information about the origin, history and transformations of data.

Total cost of ownership: The complete lifecycle cost of acquiring, operating, securing and retiring a system.

Review Questions

Short Answer Questions

1. What are the five layers of a connected healthcare IoT architecture?
2. How does remote patient monitoring differ from teleconsultation?
3. Why can excessive alerts reduce the value of a healthcare IoT system?
4. Identify three smart-city applications in utilities or infrastructure.
5. What is meant by digital inclusion?

Long Answer Questions

1. Explain how connected medical devices can improve healthcare while also creating clinical and privacy risks.
2. Analyse the organisational requirements for a successful remote patient monitoring programme.
3. Discuss how IoT can improve urban mobility, and evaluate potential equity concerns.
4. Explain the importance of interoperability, cybersecurity and governance in smart-city systems.
5. Evaluate the role of IoT in environmental management and climate resilience.

Analytical / Application Questions

1. A hospital receives 2,400 expected RPM measurements in a month, of which 2,040 are valid. Calculate the adherence rate and explain why this metric alone cannot establish clinical effectiveness.
2. Design an IoT implementation plan for detecting water leakage in a city district. Specify stakeholders, data, response processes, performance measures and key risks.
3. A municipality proposes facial-recognition cameras for general public safety. Apply necessity, proportionality, privacy, equity and accountability principles to recommend whether and how the proposal should proceed.

Multiple Choice Questions with Answers

Question 1

Which condition is essential for remote patient monitoring to create clinical value?

A. Every patient must use the same device model

B. Collected data must be connected to an accountable clinical response

C. All measurements must be continuous

D. The system must eliminate clinician involvement

Correct Answer: B. Collected data must be connected to an accountable clinical response

Question 2

A programme expects 30 measurements and receives 27 valid measurements. What is its adherence rate?

A. 80%

B. 85%

C. 90%

D. 95%

Correct Answer: C. 90%

Question 3

Which problem is most directly associated with excessive low-value healthcare alerts?

A. Data portability

B. Alarm fatigue

C. Structural corrosion

D. Demand response

Correct Answer: B. Alarm fatigue

Question 4

Which is an example of smart mobility?

A. Adaptive traffic-signal timing

B. Vaccine cold-chain monitoring

C. Bridge strain measurement

D. Hospital medication reconciliation

Correct Answer: A. Adaptive traffic-signal timing

Question 5

What is the primary purpose of a smart meter?

A. To diagnose patients remotely

B. To record and communicate resource consumption

C. To replace all utility staff

D. To provide emergency medical treatment

Correct Answer: B. To record and communicate resource consumption

Question 6

Why is interoperability important in a smart city?

A. It ensures that every system has the same supplier

B. It allows relevant systems to exchange and use information

C. It removes the need for cybersecurity

D. It guarantees that all data are accurate

Correct Answer: B. It allows relevant systems to exchange and use information

Question 7

Which measure is most relevant to evaluating digital inclusion?

A. Number of sensors installed only in affluent districts

B. Coverage and usability across different demographic groups

C. Total number of software updates

D. Average device purchase price alone

Correct Answer: B. Coverage and usability across different demographic groups

Question 8

What does structural-health monitoring observe?

A. Patient medication adherence only

B. Traffic-ticket payments

C. Conditions such as vibration, strain or corrosion in assets

D. Household internet passwords

Correct Answer: C. Conditions such as vibration, strain or corrosion in assets

Question 9

Which principle limits the collection of unnecessary personal information?

A. Purpose limitation and data minimisation

B. Vendor lock-in

C. Unlimited retention

D. Open public identification

Correct Answer: A. Purpose limitation and data minimisation

Question 10

Why should a digital twin not be treated automatically as a complete replica of a physical system?

A. It cannot contain any data

B. Its reliability depends on assumptions and the freshness and quality of source data

C. It is always operated without software

D. It is designed only for entertainment

Correct Answer: B. Its reliability depends on assumptions and the freshness and quality of source data