Sage Journals: Discover world-class research

Abstract

Objective: In the present work, LinkAll is introduced as a novel architectural model designed to enable real-time monitoring and cross-referential data analysis in remote monitoring systems across human, animal, and environmental health domains. LinkAll leverages Edge-Computing and Internet of Things principles to handle data collection, processing, and presentation from various sources. Methods: Two sibling systems were implemented to demonstrate its capability, one for monitoring urban greenery and the other for elderly home care. These systems were evaluated based on their ability to integrate with existing information systems, collect biophysical parameters, and ensure data cross-referencing. Results: Both systems demonstrate effective pluggability and cross-referenceability performances, meeting the stakeholders’ requirements. LinkAll’s ability to integrate diverse sensors and devices into existing infrastructures while providing real-time, machine-actionable insights, is also underscored. Conclusion: Pluggability, cross-referenceability, and compliance with FAIR principles make the architectural model introduced a robust solution for integrating human, animal, and environmental health monitoring systems, enhancing decision-making and contributing to One (Digital) Health’s strategic goals.

Keywords

cloud-edge computing health information exchange internet of things one health

Introduction

In the healthcare sector, the importance of assessing the evolution of patients’ conditions via remote monitoring of pre-defined set of parameters has so far increased dramatically, in terms of either health and wellbeing. Still, overall health status monitoring by means of wireless technologies is nowadays critical for either pets, livestock, and wildlife.^1,2 Further, the advances in the remote monitoring field have brought to the rise of the so-called sensor-based smart environment monitoring (SEM) systems, meant to deal from a modern perspective with phenomena such as water/air pollution and weather changes.^3,4

Human, animal, and environmental health are different yet intertwined domains, as stated by the globally acknowledged One Health concept.^5,6 Dealing with data collection and analysis in such an overarching context is therefore critical as many different sources demand for data integration and harmonization. As an instance, tackling complex issues like zoonotic diseases or environmental contaminants’ impact requires cross-referencing data from various sources (e.g., sensors for animal health monitoring and environmental sensors tracking air and water quality^7,8). Moreover, providing monitoring solutions that are pluggable into existing information systems^9,10 is vital in projects that involve collaboration among subjects, institutions, and agencies from the three Health Domains (Human/Animal/Surrounding Environment): the necessity to set their work on a collaborative, interdisciplinary approach often clashes in fact against the (sometimes remarkable) differences in terms of available data infrastructures.¹¹

With specific reference to subjects’ remote monitoring, the main challenges to be overcome for any designing solution revolve around the capability to (i) flank already existing information systems (pluggability), (ii) handle data coming from different sources (cross-referenceability), and (iii) collect continuous data from non-interacting subjects (non-invasiveness).^12,13

To address such peculiar constraints a novel conceptual framework is needed, which is able to cope with data from collection (from different domains) to presentation (to different stakeholders). In the present paper we introduce LinkAll, an architectural model that leveraging Edge-Computing (EC) and Internet of Things (IoT) principles enables real-time monitoring, and cross-referential data analysis. This architecture is mainly designed so that its implementations – the monitoring systems – can be integrated with any existing information system across various potential application fields without the need to re-engineer the latter deeply. LinkAll design allows near-source data processing enabling efficient preliminary analysis and filtering before transmitting them to the cloud. Its flexibility enables integration with existing information systems used by stakeholders with varying technological resources. Moreover, the model aims to support unified cross-referencing of collected data.

Aim of this work is to make a conceptual assessment of LinkAll to evaluate its ability to support both pluggability and remote monitoring of biophysical parameters without subjects’ interaction, while still producing cross-referenceable data, in compliance with the FAIR (Findability, Accessibility, Interoperability, and Reusability) principles.^14,15 Such evaluation will be conducted via the description and the analysis of two health-related use-cases (so–called sibling systems).^4,13

The paper is structured as follows. After the Introduction and a review of the literature, LinkAll architectural design and its sibling systems are presented. The main achievements of LinkAll are therefore analysed and discussed. Some final conclusions are eventually provided. The Business Process Model underlying LinkAll is then provided in the Appendix.

Research background

Rancea et al. examine how EC supports real-time data processing at the edge of networks, reducing latency and enabling quicker decision-making in health care systems. EC’s ability to handle diverse data streams from sensors in health care and environmental domains highlights its potential for cross-domain applications, such as monitoring human biophysical parameters alongside environmental data¹⁶; Almalki et al. look at IoT-based environmental monitoring systems with particular reference to smart city frameworks. Since real-time environmental data can be collected and analyzed for broader applications, the ability to seamlessly integrate environmental and human monitoring systems is crucial for building cross-referential platforms that can address multiple interconnected domains.¹⁷

Kazanskiy et al. focus on the development of wearable, non-invasive technologies that can monitor health parameters continuously, in both clinical and everyday settings, where real-time data collection can improve outcomes. In the same vein, Ang & Lew examine the advantages coming from the continuous data collection from multiple sources, such as wearable biosensors for humans and environmental sensors for plants.¹⁸

For what concerns criticalities coming with Interoperability, Pluggability and Cross-referencing data in Health Information Systems (HISs), the works of Sicilia & Balazote and Oyeyemi & Scott outline how standards and frameworks can be used to manage and integrate data. This allows for smoother interactions between different systems, highlighting the challenges and solutions around integrating new systems into existing infrastructures^19,20; Torab-Miandoab et al. emphasize the need for cross-platform interoperability to ensure data integration from multiple systems.²¹ Plug-and-play integration models, which enable health devices to be easily incorporated into existing systems, are discussed in Abdulrazak et al. who underscore how these models reduce the need for complex configurations.²²

On a broader prospective, Furstenau et al. highlight the challenges related to the integration of big data from multiple, heterogeneous sources in health care. A focus is made on how different types of data (e.g., clinical, sensor-based, environmental) can be managed within a unified system.²³ In addition to this, Benis et al. highlight that a vision based on the One Digital Health (ODH) framework would lead to extending existing HISs by adding new components meant to facilitate interconnectedness and interoperability.²⁴

Challenges also come from cross-domain data integration: in this regard, Zheng et al. and Hassani et al. stress on the difficulties in integrating data from various health-related domains, and wearable devices.^25,26 This is particularly true for smart cities-focused scenarios, where comprehensive frameworks are needed to manage data and support health-oriented policies to transform health ecosystems with digital technology.²⁷

Besides this, according to Nakagawa et al. when developing a concrete software architecture (SA), a (software) reference architecture (RA/SRA) is necessary to address the business rules, architectural styles, as well as best practices of software development. All of this must be supported by a unified, unambiguous, and widely understood domain terminology.²⁸ Any SA implementation process has therefore to rely on a FAIR-compliant RA that serves as a guiding blueprint that underscores the effective connections between the FAIR Principles with specific application details.²⁹ In this regard, classifications are provided in the literature^30,31 which distinguish between: (i) General Purpose Big Data SRAs, which emphasize the provision of real-time analytics to support users in decision-making, although not necessarily aligning with the FAIR Principles; (ii) Pipelines to Implement Specific FAIR Repositories to the context of specific data sharing repositories; and (iii) FAIR-compliant Big Data SRAs, as solutions concerned with the requirements imposed by the FAIR Principles, other than handling the intrinsic characteristics of big data environments.

LinkAll architecture design

Edge Cloud is fundamentally about relocating resources closer to the edge. Although all the specific architectures that leverage this technology can vary significantly, they all share a common structure consisting of three primary layers: Cloud Layer (CL), Edge Node Layer (ENL) and Edge Devices Layer (EDL).

LinkAll was designed for dedicated deployment in research projects, where the customer entity is identified as a public stakeholder institution that is often not equipped with a state-of-the-art Information System, yet requires a comprehensive data monitoring infrastructure. The main design goal was then to enable different agents, algorithms, and human staff to access and work with cross-referenceable data due to the potential multiple purposes requested in research projects. Such a goal resulted into a precise choice in the LinkAll design: a CL infrastructure deployed at stakeholder’s premises that helps in complying with data protection rules, as CL typically handles intensive processing and stores vast data volumes.

The bottommost layer of LinkAll architecture is the EDL that comprises endpoints connecting from the edge, namely the field of intervention, to request services. This layer is designed to embed low-power Personal Area Network (PAN)-connected sensors ensuring extended battery life. Such devices are needed to operate non-invasively. This means that they are meant to be set up in the system once and left alone for the entire monitoring campaign to implement a high-level non-invasiveness.

Another design choice was the adoption of Gateways.^32,33 A Gateway connects EDL and ENL to streamline network load and ensure rapid responses for time-sensitive tasks. It oversees a star network of sensors, handling data aggregation, routing, and local analysis to adapt systems’ behaviour based on sensor data and customer policies.

ENL sits between the CL and the EDL, routing communication between Gateways and Cloud Node (CN). Most of the Internet traffic occurs between Gateways and ENs. The latter can be located and replicated to fit client needs, oversee Gateways, follow CN policies, and assist in data processing without identifying subjects or storing sensitive data. This prevents information leaks and data correlation attacks.

Last layer of LinkAll is the CL that features a Data Center Network, or Cloud Node. It can handle intensive processing from different agents operating with different aims on vast data volumes, and facilitate complex operations by cross-referencing all collected data.

The mentioned aspects are summarized in the logical schema reported in Figure 1.

Figure 1.

LinkAll logical schema.

Case studies

In this section, two implementations of LinkAll will be presented. The two use cases, one from the Environmental domain and one from the Human domain, relate to two institutional stakeholders called to deal with different, complex real-life scenarios. The systems are referred to as “sibling” because they were conceived, developed, and deployed directly upon the pre-existing architectural model, thus sharing its fundamentals. The differences between the sibling systems are basically related to the differences in the fields of application. This leads in turn to different employed hardware and customizations requested by the stakeholders.

Hardware and software deployment summary

In Table 1 the main components of LinkAll implementations and their role in the architecture are described.

Table 1.

Main components of LinkAll implementations and their role in the architecture.

Component	Type	Role in architecture	Deployment count
Environmental sensors	BLE-based (soil temp, moisture)	EDL	100 (urban greenery)
Vital monitoring devices	BLE/Wi-Fi (e.g., SpO₂, ECG, motion sensors)	EDL	250 (elderly care)
Gateway units	ESP32/Android set-top box	Data transport & policy exchange between EDL and ENL	5 (urban)/50 (home)
Connectivity modules	GSM/Wi-Fi chipsets	Edge and cloud data transmission	Integrated in gateways
Cloud node instance	Hosted VPS (Linux based)	Central orchestration & Policy control	1 (for each domain)
Edge node instance	Hosted VPS (Linux based)	Data routing & Policy deployment	2 (urban)/1 (home)
DB & services	PostgreSQL	Data ingestion, storage, and event processing	Included in CN
Security layer	NGINX + TLS certs + Auth tokens	Secure access & data protection	Included in each EN
Semantic broker	Custom JSON-LD transformer + Ontology mapper	Metadata harmonization & FAIR compliance	Included in each EN

A deployment diagram with more technical details is also provided in Figure 2.

Figure 2.

LinkAll deployment diagram.

Figure 3 emphasizes the sequence by which stakeholders’ requests are encoded into policies, which are enforced via specific gateways configurations and then turned into commands for devices.

Figure 3.

Request-policies-configuration-commands sequences in sibling systems.

Remote monitoring system for urban greenery

It is currently of utmost importance to design IoT–based ecosystems to support data and services integration in complex environments, such as green areas within smart cities, to positively impact on human and animal health and wellness.³⁴ To this end, a collaborative research project with the Municipal Administration of a Southern Italian Metropolitan city led to the development of a sophisticated remote monitoring system for urban greenery. The system enables specialists to evaluate damage from severe weather, environmental stress, or human activity by capturing and processing biophysical data—including wind speed, air and soil temperatures, humidity, and the motion and orientation of plants. Data are then uploaded to a cloud-based eGovernment platform for analysis and response, which represents the CL of the LinkAll architectural model.

The project involved the non-intrusive placement of sensors on plants across a large open-air area, requiring elevated and root-level installations. Due to their challenging placement on greenery, these outdoor sensors needed to be durable, cost-effective, and to require low maintenance. Green computing methods³⁵ were applied to optimize power use and extend battery life while ensuring sufficient wireless range to enhance their longevity and efficiency. The sensors were housed in water- and dust-resistant casings to protect against environmental exposure. They also used a System-on-a-Chip (SoC) with Bluetooth Low Energy (BLE) support, extending yearly maintenance cycles. Additionally, the sensors featured remotely adjustable signal amplification to effectively balance power consumption with transmission range.^36,37 These IoT devices belong to the low power PAN-connected sensors family and can then be set up and left in place to work for the entire monitoring campaign without maintenance.

A custom Gateway Station was developed making use of an ESP-32-WROVER module microcontroller with a GSM modem for internet connectivity.³⁸ The Gateway Station stands in between the ENL and the EDL. It therefore handles tasks such as receiving configuration updates from the Edge Node, adjusting sensor settings, collecting sensor data, and transmitting them back. When new sensors are added to the monitoring system, they are dynamically integrated into the closest Gateway, which is promptly reconfigured. Similarly, sensors are reassigned to the closest Gateway to optimize connectivity, if a new one is installed. In addition, the Gateway compresses data and activates the GSM modem only as needed to conserve power; the Station also collects local weather information, providing measurement context.

The ENL combines consumer electronics with open-source software and handles data operations and Gateway management. It features a REST interface for communication with the CL. The adoption of Representational State Transfer Application Programming Interfaces (REST APIs) was mainly due to REST’s intrinsic flexibility, which implies a better suitability to more innovative contexts, such as IoT and mobile applications.³⁹ Data processing starts with decompressing data units and then processing measurements according to sensors’ installation parameters stored in the EN. This helps deducing time series data about the inclinations of branches and masts. Data are ultimately transmitted to the CN. Gateway management dynamically generates configurations that minimize power use and incorporate policies from the CL, which is part of the Municipal Administration’s Information System.

The CN receives data from the EN and exhibits a graphical interface for either sensor and policy declarations and administration, and for data visualization for domain experts. Under standard policies, the CN reports data like wind speed, acceleration, temperature, and humidity at varying frequencies, based on each parameter’s importance. As an instance, branch acceleration data updates occur more frequently than trunk data, and temperature and humidity change over extended periods. This allows for sensor adjustments to optimize battery life and reduce data redundancy. Additionally, policies can be tailored to specific scenarios to enhance precision forestry interventions.⁴⁰ It must be underlined that the CN has not been entirely developed by the authors, who instead had in charge the integration of the Cloud components needed for LinkAll operations with the Municipal Administration Information System, while paying attention to the existing effective policies.

Elderly individuals home care

An effective monitoring of (bio)physical parameters is also critical under a medical–social intervention point of view, for maintaining a good health-related quality of life.^41–43 In this regard, a collaborative effort with a University Hospital devised a scenario involving elderly individuals participating in a home care program incorporating Assisted Physical Activity for a telerehabilitation prototype.⁴⁴ The participants were provided with medical-graded IoT devices to conduct exercises and conveniently monitor biophysical parameters. The data generated from these activities were seamlessly gathered, analyzed, and incorporated into patients’ Electronic Medical Records (EMR) centrally managed on the hospital’s CN.

This research project, named EASYDOM by Multiplate AGE, was granted ethical approval by the A.O.U. Federico II – A.O.R.N. “Antonio Cardarelli” Ethics Committee (protocol nr. 213/21). The collected and processed data during this research are stored in a secured, encrypted manner, with restricted access provided by A.O.U. Federico II - Department of Clinical Medicine and Surgery.^45,46

The sensors employed in the project fall into two categories:

• The first one is called “set-and-forget” because a device just needs to be paired with the Gateway at the first installation of the system. This provides the connection and the data download tasks without any patient interaction for the entire monitoring campaign. The devices considered are e.g., fitness trackers, speed sensors in mini-exercise bikes, and inertial sensors for monitoring heart rate and sleep on bedding. In this category only fall devices designed and produced to be used in fitness environments with a personal monitor aim;

• The second one features interactive devices such as thermometers, pulse oximeters, blood pressure monitors, and scales. In this category only fall devices designed to produce medical graded measurements, so they can be used either by patients at their premises, or by medical personnel at healthcare structures.

To make sensors from the latter category capable of operating similarly to their smart counterparts (from the former category) an ad-hoc devices integration was implemented. As a result, patients were simply required to start a measurement while the Gateway handled any setup or connectivity, regardless of the device used. This allowed therefore all of the monitoring devices embedded in this campaign to enter fully entitled into the EDL.

Budget-friendly appliances running Android OS on an ARM-based SoC serve as Gateways. The choice of Android OS was sort of mandatory since smart devices’ producers only supplied APIs designed for this operating system. An in-house application running on the gateways facilitates BLE sensor communication for data management, including measurements and events like devices moving out of range. Behavioral policies are set by the EN, allowing configurations updates during operation without user intervention, such as defining new data targets or altering operation schedules and security credentials. Each Gateway transmits collected data to the ENL for further processing. It can also make fundamental decisions based on sensor readings, like triggering notifications on fitness trackers. Installation requires power and Local Area Network (LAN) cable connection to the patient’s router for internet access. The use of a cabled connection prevents patients from the burden of connecting gateways to I/O devices to apply the local network configuration. At startup, each Gateway fetches its configuration from the deputed EN to recognize connected sensors, thus forming a combined kit comprising both the gateway and the sensors.

The ENL, powered by consumer electronics-grade computational power and open-source software, addresses a manifold task: (i) communicates with Gateways through a message queue-based interface, (ii) provides a REST service for interactions with the CL, (iii) processes data from gateways, handling tasks like unit conversion, filtering, aggregation, and compression, (iv) updates gateway configurations and schedules based on policies received from the CL via the REST interface, influencing sensors’ fleet composition and sampling settings, and (v) stores sensor identifiers, aiding in configuring gateways and defining their operational behaviors.

The CL receives data from the ENs and provides interfaces for medical staff, algorithms, and data visualization. However, stakeholders are responsible for partially implementing and fully commissioning CL, since it primarily hosts the hospital EMR system. Authors were responsible of developing all the components needed for the integration of the system into the EMR already in use by the stakeholder, in accordance with the LinkAll architectural model. This setup enables medical staff to access a unified interface, consolidating patient data from local sources and Edge Devices. When a new patient is enrolled into the remote monitoring plan, medical staff register their data on the platform and link them with a gateway and related sensors, guiding patients on sensors’ usage without requiring any further configuration or maintenance beyond essential battery management.

As the monitoring system operates, Gateways continuously receive measurements from various Edge Devices, collecting data concurrently from all patient-used devices. For instance, weight data from a scale and temperature from a thermometer can be gathered simultaneously, facilitated by BLE protocol performance. Some Edge Devices, like smartwatches and portable ECGs, store multiple parameters for several days in built-in memory. When the monitoring devices are within a gateway range, this retrieves and processes the entire device memory content to eliminate redundancy and update the Edge Node with new data.

Discussion

The LinkAll architecture is not a generic IoT–Cloud stack. It represents a domain-agnostic, policy-driven IoT–Edge–Cloud framework purpose-built to deal (and connect) with different Health-related domains. Unlike conventional IoT infrastructures that are domain-specific, tightly coupled, or require manual configuration, LinkAll is characterized by three distinctive features:

• Runtime policy-based reconfigurability of Edge Devices and Gateways without firmware reflashing;

• Layered architectural separation (Edge Devices Layer, Edge Node Layer, Cloud Node) with embedded data governance at each level;

• Semantic cross-referenceability, using metadata harmonization and shared ontologies to enable FAIR-compliant data fusion across human, animal, and environmental domains.

This design departs from open IoT platforms such as FIWARE, which provide broker-centric architectures but lack native semantic control and policy orchestration.^47–49 LinkAll’s novelty lies not in the invention of new protocols, but in the architectural integration of existing standards. In other words, the proposed architecture adopts a modular structure that supports a progressive approach to the complex issue of interfacing different systems from different domains without the need of altering them (non-invasiveness). This makes it possible to tackle One Health–related challenges and to achieve related strategic outcomes in complex real-life scenarios, thus aligning with an ODH reference model that supports intervention-level governance and federated deployments.^27,50–53

The use cases were designed from the ground up to validate the flexibility of LinkAll in real-life scenarios across structurally different domains, involving distinct sensor classes, stakeholder ecosystems, and operational requirements:

• The urban greenery system relied on outdoor, GSM-based gateways, enabled by LinkAll’s adaptive data buffering and policy-driven reconfiguration to react to the stakeholder’s needs effectively (e.g., sampling frequency based on seasonal exposure);

• The elderly home care system utilized heterogeneous BLE and Wi-Fi medical sensors, where new policies were dynamically applied through the CN’s orchestration layer.

System performance evaluation

To provide an insight as to performance-related aspects of the LinkAll architecture, a targeted empirical evaluation was conducted in both implemented use cases over a 10-day period. System logs were collected directly from the CN and the ENL, using embedded monitoring scripts that recorded timestamped events, transmission cycles, gateway-level buffering behavior, and success/failure packet counters. Power consumption was estimated via current profiling utilities available on embedded boards, combined with device-specific manufacturer documentation.^54,55 The metrics collected are summarized in Table 2.

Table 2.

System performance metrics.

Metric	Urban greenery use case	Elderly home care use case	Measurement approach
Avg. latency (ms)	610 ± 125	415 ± 85	Timestamp delta: data creation (EDL) → ingestion (CN); N = 1000 data packets
Throughput (packets/min)	∼170	∼240	Recorded at ENL queue over 60-min intervals per gateway
Data reliability (%)	98.7%	99.4%	(Total delivered packets/total sent) × 100 over 10-days logging window
Packet loss (per gateway)	2.8 packets/day	1.2 packets/day	Differential logging between EDL → ENL → CN
Avg. sensor power consumption (mW)	78 mW (PAN-connected BLE)	66 mW (BLE & Wi-Fi mixed devices)	Estimated via onboard logs + BLE module spec sheets + duty cycle profiling
TLS handshake time	140 ms (±20 ms)	180 ms (±20 ms)	HTTPS REST calls with connection reuse; N = 50 attempts

The average Gateway-to-Cloud latency was assessed by using 128-byte sensor payloads under both idle (1–2 devices active) and peak (simulated 15–20 concurrent devices) conditions. The values achieved are consistent with benchmarks for similar IoT–Edge–Cloud architectures for non-critical but responsive applications.^56,57 The system’s throughput remained stable even under moderate transmission bursts, supported by the use of adaptive buffering at the Gateway level and on-demand GSM/Wi-Fi connectivity policies. Notably, reliability exceeded 98% across both use cases, thus aligning with targets suggested by many scholars for cross-domain EC deployments in smart cities and healthcare applications.^37,58 The packet loss observed was largely attributable to temporary gateway unavailability or GSM signal dropouts in the urban greenery case.

From a power efficiency standpoint, the PAN-connected low-energy sensors in the urban case required less than 80 mW on average. These figures align with continuous duty cycles at 10–60 s sampling intervals and are comparable to values reported in the literature for non-intrusive IoT deployments.^18,59 Transport Layer Security (TLS) was selectively enabled for Gateway–CN communication via RESTful HTTPS endpoints. The Connection Reuse was enabled, reducing per-request cost post-handshake, while the Data Upload Latency Impact was equal to +9%–11% of total transaction time on first request, <5% average thereafter.

Architectural flexibility and scalability

The LinkAll architecture was explicitly designed for modularity and scalability. Its flexibility was validated through the two structurally different sibling systems analyzed, demonstrating adaptability to both environmental and clinical domains without altering core architectural logic.

At the EDL, the system leverages open protocols (BLE, Zigbee, Wi-Fi) and adopts a loosely coupled communication strategy that supports hardware heterogeneity and device substitution. This design aligns with recommendations on emphasizing that scalable IoT–Edge–Cloud frameworks must allow seamless onboarding of diverse sensor types without extensive reconfiguration.^33,60 Moreover, device-level flexibility is enhanced by a policy-driven abstraction layer at the ENL, capable of interpreting hardware-specific data streams through dynamically loaded templates, as also proposed for Smart City platforms.^29,61

At the ENL, scalability is achieved via microservice-inspired functional decomposition, where Gateway tasks – data parsing, compression, routing, and diagnostics – are containerized and governed by configuration policies received from the CN. This modularity enables either vertical (e.g., supporting an increase in sampling frequency or data volume) and horizontal scalability (e.g., adding Gateways or ENs). Use cases implementation showed stable performance when scaling from 5 to 30 devices per gateway, consistent with architectural expectations described in the literature.^32,62

At the CL, REST-based service orchestration allows seamless integration of external systems (e.g., EMRs, city management systems, weather APIs) while maintaining independent control over data storage, processing, and access policies. This facilitates federation across geographically and institutionally distributed deployments, thus echoing what already stated for cross-domain health informatics infrastructures.^26,63

Moreover, LinkAll’s policy-based orchestration allows remote reconfiguration of sensors and Gateways without physical access or software redeployment. For instance, in the urban greenery use case, changes in weather-driven risk profiles triggered updates in sampling resolution and upload frequency in real time. Such autonomous behavior based on contextual triggers is increasingly seen as a requirement for scalable health IoT systems.^22,23

The architecture also accommodates deployment in both dense urban environments (via Wi-Fi mesh or fiber-backed Gateways) and low-connectivity rural areas (via GSM and buffered store-and-forward logic), addressing a common scalability challenge in distributed public health systems.^34,64

LinkAll positioning in remote monitoring scenario

A common objective in projects addressing human, animal, and environmental health is to provide solutions tailored to institutional stakeholders. These solutions should integrate with existing Information Systems and facilitate cross-referencing across diverse data sources. Ensuring the accuracy and reliability of measurements via the embedment of certified measurement devices is crucial as well. For such reasons the ensemble of the FAIR–compliant LinkAll key features is hard to be found in other existing solutions, be they either market solutions or developed through applied research.

Solutions from large vendors often lack features related to pluggability, primarily for commercial reasons. Minimal interest stands in creating solutions that can integrate with other systems, as the goal is to sell a closed, single-package (although customizable) solution to the end customer. It’s highly likely that features related to referenceability are implemented, as it is often observed that different devices from the same vendor display aggregated information via a single, shareable interface, such as websites or smartphone apps. Furthermore, as large vendors strive to manufacture and market certified devices, the accuracy and reliability of the measurements are typically top-notch – with specific reference to the medical devices market, it is the case of corporations such as Omron, Withings, or Beurer.⁶⁵ While these vendors excel at creating closed ecosystems with seamless internal integration, the challenges arise when trying to extend pluggability and cross-referenceability across external systems or third-party devices. Moreover, while these devices provide reasonably accurate measurements for consumer use, achieving clinical-grade accuracy consistently remains a challenge: overcoming these complexities requires the adoption of industry-wide standards, enhanced data interoperability protocols, and the use of advanced algorithms to ensure accuracy and seamless integration across different platforms.^66,67

Solutions developed in contexts focused on applied research can include features related to pluggability and referenceability. The devices embedded in these solutions are primarily experimental and designed for rapid prototyping, making them highly customizable: the sensors integrated in such solutions are in fact often “homemade” creations, developed using corresponding microcontroller modules. This allows high-level results to be achieved for both of the mentioned architectural features. However, the significant focus on the simplicity of device integration and the ease of platform creation may hamper the timely development of measurement accuracy and reliability, thus resulting in largely inaccurate readings, as the sensors are more experimental prototypes than fully engineered, market-ready devices. These modules are primarily intended for use in software architecture design and need to be replaced with proper monitoring devices that incorporate the corresponding sensors for reliable, real-world applications.^68,69

LinkAll can fulfil the need to use data collected across scientific disciplines and related to different communities, as the proposed architecture has been designed with a focus on cross-referenceability. Data collection from the whole set of Edge Devices is ultimately achieved by exploiting low-level communication protocols, thus giving the authors complete control of the (multi-terminology) thesaurus to be specifically designed.⁷⁰ In this way, data from different sources across different Health domains can essentially share a common vocabulary and background knowledge. The intentional creation of an ecosystem that facilitates seamless and secure health data exchange and processing aligns with the requirements of Interoperability. At the semantic level, the terminology used for (meta)data, the definition of variables, and their attributes have been designed to align seamlessly with a controlled vocabulary such as an ontology.⁷¹ In fact, many project stakeholders requested data collection applications with low-cost, non-invasive sensors, which they could use to study, process, and cross-reference the data using their existing Information Systems.⁷² An application that leverages the IoT and EC principles can be connected to these, allowing the principles to complement each other in a syndemic view.

Further, incorporating characteristics of pluggability makes the adoption of the solution easier and less disruptive for stakeholders within a real-world scenario, either for different application contexts within the same domain, or in different but interconnected domains. Pluggability for LinkAll can also be intended in a wider sense: thanks to the design choices made it is possible to integrate - almost - any kind of monitoring devices in the deployed systems, as long as devices’ communication protocols are available or commonly known. This makes it theoretically possible to plug diverse monitoring devices, even entire monitoring solutions, in a system that implements LinkAll architecture, which shows then as a totally scalable and highly versatile framework. To push towards the pluggability limits, design choices were also made to produce data featuring as rich as possible attributes. These include provenance information and globally unique identifiers, which improve data traceability. A systematic, ongoing, and intelligent integration of extensive, multidimensional data exchanges was therefore enabled, thus aligning with the requirements of Findability and Reusability.

Achieving results in terms of accuracy and reliability of measurements was challenging. The main issues related in fact to finding various types of certified sensors capable of producing reliable and accurate medical-graded measurements and enabling them to connect to a single gateway. This often implied requesting the manufacturer to provide either the protocol documentation or a library implementation of the communication interface. The sensors commonly used to sample both medical/non-medical data rely on BLE^73,74 because they do not require substantial throughput but rather connectivity that backups batteries. As BLE specifications dictate hardware and connectivity protocols but not the content of the transmitted packets, many sensors use proprietary data protocols and representations, necessitating proprietary programs or smartphone applications to access the data. Since these constraints would not match the requirements of Accessibility, the research group worked toward the complete integration of each sensing device into the sibling systems.

Table 3 provides a summary of the discussed results.

Table 3.

Features supported by different kinds of solutions (✓: supported; X: unsupported).

	Market vendors		Applied research		LinkAll
Pluggability	X	Industry-wide standards, enhanced data interoperability protocols requested toward an extension across external systems or third-party devices^88,89	✓	Highly customizable experimental devices, designed for rapid prototyping. Significant focus on the simplicity of device integration and the ease of platform creation^90,91	✓	Possibility to integrate - almost - any kind of monitoring device in the deployed systems, provided that devices’ communication protocols are available or commonly known⁷⁰
Cross-referenceability	✓	Different devices from the same vendor display aggregated information via a single, shareable interface^13,66,92	✓		✓	Exploiting low-level communication protocols allows for complete control of a (multi-terminology) thesaurus⁷¹
Accuracy and reliability of measures	✓	Closed ecosystems with seamless internal integration^67,93	X	Experimental prototypes (not yet fully engineered) can be affected by inaccurate readings^69,94	✓	Efforts made toward the complete integration of each sensing device, both medical-graded and non, into the sibling systems⁷²

Limitations and future research directions

While the LinkAll framework offers significant advantages, its implementation in real-world scenarios presents either several practical limitations, and room for future research. These limitations do not directly concern flaws in the framework itself but rather challenges that arise from the complexities of integrating diverse technologies and operating within dynamic environments.

Firstly, significant challenges come from regulatory constraints, especially when dealing with medical-grade devices and patient data. The strict regulations governing data privacy (e.g., GDPR, HIPAA) and medical device certification (e.g., FDA, CE marking) necessitate careful consideration during design and deployment.^75,76 While LinkAll aims to integrate certified devices and ensure secure data handling, the processing of data at the gateway or EN level could potentially affect the regulatory compliance of measurements if not handled with extreme care. Future work will focus on developing clear guidelines and mechanisms to ensure that any edge processing maintains the integrity and regulatory compliance of medical device data. However, data is currently incapsulated in structures and not manipulated during the whole process, so regulatory compliance is well preserved.

Secondly, BLE congestion can become a limiting factor in dense urban deployments or environments with a high concentration of BLE devices. While BLE is energy-efficient and suitable for many IoT applications, its limited bandwidth and potential for interference can lead to packet loss and increased latency, especially when numerous devices are transmitting data simultaneously.⁷⁷ This could impact the real-time performance and reliability of LinkAll in scenarios with a large number of co-located sensors. Future research will explore advanced scheduling algorithms, alternative low-power communication protocols, and dynamic frequency hopping techniques to mitigate BLE congestion.

Thirdly, intermittent connectivity is a common challenge in remote monitoring scenarios, particularly in geographically dispersed or mobile deployments. While the urban greenery system utilizes GSM for connectivity, and the elderly home care system relies on LAN, disruptions in internet access can temporarily halt data transmission to the CL. Although ENs can perform local processing and temporary storage, prolonged disconnections can lead to data synchronization issues and delays in critical insights.⁷⁸ Future work will investigate robust offline data synchronization mechanisms, opportunistic networking strategies, and more resilient communication protocols to enhance LinkAll’s performance in environments with unreliable connectivity.

Finally, projections can be made regarding efficiency and hardware demands. For instance, in terms of system scalability, it will be necessary to provide different, more powerful Gateways to maintain high efficiency. Regarding privacy policies, different cryptographic algorithms may need to be implemented depending on the country or regulation, further adding complexity to the system if multiple regions with distinct policies are to be served simultaneously.⁷⁹ Challenges are anyway always likely to arise as of balancing cost and effectiveness of the solutions to be designed.

Conclusions

In this work LinkAll has been introduced as an novel conceptual framework that leverages Edge-Computing and IoT principles to enable real-time monitoring, and cross-referential data analysis. The framework’s key features - pluggability, cross-referenceability, and accuracy and reliability of measurements - were specifically developed to address challenges arising from operating in intertwining Health digital domains. LinkAll also aims to support unified cross-referencing of collected data in compliance with the FAIR principles. Two use cases were presented to demonstrate the opportunities arising with the model’s deployment in both human- and environment-centred scenarios, where stakeholders may operate non-bleeding-edge Information Systems while navigating stringent and complex privacy policies. However, access to cross-referenceable data from different domains, gathered by certified and commercially available sensors, empowers stakeholders to conduct sophisticated and precise data analysis using large datasets. Such data becomes increasingly crucial when considering AI-based approaches, signalling a promising direction for future developments in One (Digital) Health-focused scenarios.⁸⁰

Footnotes

ORCID iDs

Oscar Tamburis

Arriel Benis

Ethical considerations

[…] This research project, named EASYDOM by Multiplate AGE, was granted ethical approval by the A.O.U. Federico II – A.O.R.N. Cardarelli Ethics Committee (protocol nr. 213/21). The collected and processed data during this research are stored in a secured, encrypted manner, with restricted access provided by A.O.U. Federico II - Department of Clinical Medicine and Surgery […].

Consent to participate

Participants (or their legal representatives) to the research project, named EASYDOM by Multiplate AGE, have been informed about the study’s purpose, procedures, risks, and benefits, and have voluntarily agreed to participate.

Author contributions

OT: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Project administration, Funding acquisition.

AT: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision.

GP: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization.

GI: Formal analysis, Writing - Review & Editing, Visualization.

AB: Formal analysis, Writing - Review & Editing, Visualization, Supervision, Funding acquisition.

MM: Conceptualization, Methodology, Formal analysis, Investigation, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was partially funded by the European Union’s Horizon Europe Research and Innovation Program (HORIZON-CL6-2022-GOVERNANCE-01) under grant agreement No. 101086521 - OneAquaHealth (Protecting urban aquatic ecosystems to promote One Health).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data are not publicly available due to ethical and legal restrictions.*

Conference submission

The present manuscript is a (widely) enlarged and revised version of the following [ref. #]: Tamburis O, Tramontano A, Perillo G, et al. All for One, All at Once: A Pluggable and Referenceable Architecture for Monitoring Biophysical Parameters Across Intertwined Domains. In: Barolli L (ed) Advanced Information Networking and Applications. Cham: Springer Nature Switzerland, 2024, pp. 264–276.

Appendix

References

Jukan

Masip-Bruin

Amla

. Smart computing and sensing technologies for animal welfare: a systematic review. ACM Comput Surv 2018; 50: 1–27.

Junior

. IoT applications for monitoring companion animals: a systematic literature review. In: 2020 14th International Conference on Innovations in Information Technology (IIT), Al Ain, UAE, 17–18 November 2020. IEEE, pp. 239–246.

Dhaka

Kundu

Rani

, et al. Role of internet of things and deep learning techniques in plant disease detection and classification: a focused review. Sensors 2023; 23: 7877.

Tamburis

Magliulo

Tramontano

, et al. Precision grazing: when agriculture, livestock and technology unite. In: 2023 IEEE international conference on big data (BigData), Sorrento, Italy, 15–18 December 2023, pp. 3510–3515. IEEE.

Lerner

Berg

. The concept of health in one health and some practical implications for research and education: what is one health? Infect Ecol Epidemiol 2015; 5: 25300.

Destoumieux-Garzón

Mavingui

Boetsch

, et al. The one health concept: 10 years old and a long road ahead. Front Vet Sci 2018; 5: 14.

El Morr

Ozdemir

Asdaah

, et al. AI-based epidemic and pandemic early warning systems: a systematic scoping review. Health Inf J 2024; 30: 14604582241275844.

Zhou

Liu

Carrion

, et al. Spectro-environmental factors integrated ensemble learning for urban river network water quality remote sensing. Water Res 2024; 267: 122544.

Barros

Duggan

Rebholz-Schuhmann

. The application of internet-based sources for public health surveillance (infoveillance): systematic review. J Med Internet Res 2020; 22: e13680.

10.

Chiou

Voegeli

Wilhelm

, et al. The future of infodemic surveillance as public health surveillance. Emerg Infect Dis 2022; 28: S121–S128.

11.

Hitziger

Berezowski

Dürr

, et al. System thinking and citizen participation is still missing in one health initiatives – lessons from fifteen evaluations. Front Public Health 2021; 9: 653398.

12.

Ghosh

Chakraborty

, et al. UltraSense: a non-intrusive approach for human activity identification using heterogeneous ultrasonic sensor grid for smart home environment. J Ambient Intell Hum Comput 2023; 14: 15809–15830.

13.

Tamburis

Tramontano

Perillo

, et al. All for one, all at once: a pluggable and referenceable architecture for monitoring biophysical parameters across intertwined domains. In: Barolli

(ed) Advanced information networking and applications. Switzerland: Springer Nature, 2024, pp. 264–276.

14.

Wilkinson

Dumontier

Aalbersberg

, et al. The FAIR guiding principles for scientific data management and stewardship. Sci Data 2016; 3: 160018.

15.

Tamburis

Benis

. One digital health for more FAIRness. Methods Inf Med 2022; 61: e116–e124.

16.

Rancea

Anghel

Cioara

. Edge computing in healthcare: innovations, opportunities, and challenges. Future Internet 2024; 16: 329.

17.

FarisA

Alsamhi

Sahal

, et al. Green IoT for eco-friendly and sustainable smart cities: future directions and opportunities. Mobile Network Appl 2023; 28: 178–202.

18.

Kazanskiy

Khonina

Butt

. A review on flexible wearables – recent developments in non-invasive continuous health monitoring. Sensor Actuator Phys 2024; 366: 114993.

19.

Sicilia

MÁ

Balazote

(eds) Interoperability in healthcare information systems: standards, management, and technology. IGI Global, 2013. DOI: 10.4018/978-1-4666-3000-0.

20.

Oyeyemi

Scott

. Interoperability in health and social care: organisational issues are the biggest challenge. BMJ Health Care Inform 2018; 25: 196–197.

21.

Torab-Miandoab

Samad-Soltani

Jodati

, et al. Interoperability of heterogeneous health information systems: a systematic literature review. BMC Med Inform Decis Mak 2023; 23: 18.

22.

Abdulrazak

Paul

Maraoui

, et al. IoT architecture with plug and play for fast deployment and system reliability: AMI platform. In: Aloulou

Abdulrazak

De Marassé-Enouf

, et al (eds) Participative Urban Health and Healthy Aging in the Age of AI. Springer International Publishing, 2022, pp. 43–57.

23.

Furstenau

Leivas

Sott

, et al. Big data in healthcare: conceptual network structure, key challenges and opportunities. Digital Communications and Networks 2023; 9: 856–868.

24.

Benis

Tamburis

Chronaki

Moen

. One digital health: promoting innovation in health information systems. In Research Handbook on Health Information Systems 2025 . Edward Elgar Publishing, Feb, 10, pp. 277–293.

25.

Zheng

. Methodologies for cross-domain data fusion: an overview. IEEE Trans Big Data 2015; 1: 16–34.

26.

Hassani

Dackermann

Mousavi

, et al. A systematic review of data fusion techniques for optimized structural health monitoring. Inf Fusion 2024; 103: 102136.

27.

Benis

Haghi

Deserno

, et al. One digital health intervention for monitoring human and animal welfare in smart cities: viewpoint and use case. JMIR Med Inform 2023; 11: e43871.

28.

Nakagawa

Oliveira Antonino

Becker

. Reference architecture and product line architecture: a subtle but critical difference. In: Crnkovic

Gruhn

Book

(eds) Software Architecture. Springer Berlin Heidelberg, 2011, pp. 207–211.

29.

Santana

EFZ

Chaves

Gerosa

, et al. Software platforms for smart cities: concepts, requirements, challenges, and a unified reference architecture. ACM Comput Surv 2018; 50: 1–37.

30.

Castro

JPC

Aguiar

. Big data architectures for FAIR-compliant repositories: a systematic review. In: Anais do XXXVIII Simpósio Brasileiro de Banco de Dados (SBBD 2023). Sociedade Brasileira de Computação - SBC. 2023; 76–88.

31.

Buccella

Cechich

. An SPL reference model based on domain taxonomies & standards. JCS&T 2022; 22: e09.

32.

Martin Fernandez

Diaz Rodriguez

Rubio Munoz

. An edge computing architecture in the internet of things. In: 2018 IEEE 21st International Symposium on Real-Time Distributed Computing (ISORC), Singapore, 29–31 May 2018, pp. 99–102. IEEE.

33.

Taivalsaari

Mikkonen

Pautasso

. Towards seamless IoT device-edge-cloud continuum: software architecture options of IoT devices revisited. In: Bakaev

I-Y

Mrissa

, et al (eds) ICWE 2021 Workshops. Springer International Publishing, 2021, pp. 82–98.

34.

Javed

Kubler

Malhi

, et al. bIoTope: building an IoT open innovation ecosystem for smart cities. IEEE Access 2020; 8: 224318–224342.

35.

Benis

Tamburis

. The need for green and responsible medical informatics and digital health: looking forward with one digital health. Yearb Med Inform 2023; 32: 7–9.

36.

Tramontano

Tamburis

Magliulo

. To the green from the Bl(u)e: an innovative system for monitoring urban green areas. In: 2022 IEEE international conference on metrology for extended reality, artificial intelligence and neural engineering (MetroXRAINE), Rome, Italy, 26–28 October 2022, pp. 646–650. IEEE.

37.

Tramontano

Perillo

Magliulo

, et al. Scaling up environmental governance in precision forestry. In: Stoicu-Tivadar

Benis

Deserno

, et al (eds) Studies in Health Technology and Informatics. IOS Press, 2024. doi:10.3233/SHTI241055.

38.

ESP-WROVER-KIT V4 . 1 getting started guide - ESP32 - — ESP-IDF programming guide latest documentation. https://docs.espressif.com/projects/esp-idf/en/latest/esp32/hw-reference/esp32/get-started-wrover-kit.html# (2024, accessed 9 May 2024).

39.

Zappatore

Longo

Martella

, et al. Semantic models for IoT sensing to infer environment–wellness relationships. Future Gener Comput Syst 2023; 140: 1–17.

40.

Holopainen

Vastaranta

Hyyppä

. Outlook for the next generation’s precision forestry in Finland. Forests 2014; 5: 1682–1694.

41.

Tramontano

Feoli

Tamburis

, et al. Deploying unsupervised learning for daily activity windows analysis in cancer patients. In: 2023 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE), Milano, Italy, 25–27 October 2023, pp. 1188–1192. IEEE.

42.

Dobre

Băjenaru

Drăghici

, et al. Sustainable health-related quality of life in older adults as supported by the vINCI technology. Sensors 2023; 23: 2287.

43.

Barillaro

Feoli

Tramontano

Comerci

Caroprese

Cuocolo

Tamburis

Petrazzuoli

D’Arienzo

Farella

Oliviero

. Fatigue trajectories by wearable remote monitoring of breast cancer patients during radiotherapy. Scientific Reports. 2024 Nov 814;(1):27276.

44.

Benis

Banker

Pinkasovich

, et al. Reasons for utilizing telemedicine during and after the COVID-19 pandemic: an internet-based international study. JCM 2021; 10: 5519.

45.

Pilotto

Barbagelata

Morganti

, et al. Development and implementation of multicomponent homecare interventions for multimorbid and frail older people based on information and communication technologies: the MULTIPLAT_AGE project. Arch Gerontol Geriatr 2024; 117: 105252.

46.

Anziani ed attività fisica adattata: parte il progetto EASYDOM by MULTIPLATE AGE. https://www.policlinico.unina.it/flex/cm/pages/ServeBLOB.php/L/IT/IDPagina/3917 (accessed 9 May 2024).

47.

Montori

Liao

De Giosa

, et al. A metadata-assisted cascading ensemble classification framework for automatic annotation of open IoT data. IEEE Internet Things J 2023; 10: 13401–13413.

48.

FIWARE – foundation. https://www.fiware.org/foundation/ (2017, accessed 17 July 2025).

49.

IFIP Advances in Information and Communication Technology. Exploring FIWARE-based data model for data sharing platforms. Switzerland: Springer Nature, 2025, pp. 126–139.

50.

Haghi

Benis

Deserno

. Accident & emergency informatics and one digital health. Yearb Med Inform 2022; 31: 40–46.

51.

Tsai

Mendenhall

Trostle

, et al. Co-occurring epidemics, syndemics, and population health. Lancet 2017; 389: 978–982.

52.

Benis

Notea

Barkan

. Risk and disaster management: from planning and expertise to smart, intelligent, and adaptive systems. Stud Health Technol Inf 2018; 247: 286–290.

53.

Benis

Tamburis

Chronaki

, et al. One digital health: a unified framework for future health ecosystems. J Med Internet Res 2021; 23: e22189.

54.

Putra

MAP

Hermawan

Kim

D-S

, et al. Data prediction-based energy-efficient architecture for industrial IoT. IEEE Sens J 2023; 23: 15856–15866.

55.

Raghavendar

Batra

Malik

. A robust resource allocation model for optimizing data skew and consumption rate in cloud-based IoT environments. Decision Analytics Journal 2023; 7: 100200.

56.

Towards characterization of edge-cloud continuum. In: Lecture Notes in Computer Science. Springer International Publishing, pp. 215–230.

57.

Truong

Pham

, et al. Developing a BI-based framework for assessing the smart levels of hospitals. Health Inf J; 31: 14604582251353546.

58.

Bibri

Krogstie

Kaboli

, et al. Smarter eco-cities and their leading-edge artificial intelligence of things solutions for environmental sustainability: a comprehensive systematic review. Environ Sci Ecotechnol 2024; 19: 100330.

59.

Pan

Zhang

, et al. ACKnowledge: a computational framework for human compatible affordance-based interaction planning in real-world contexts. In: Proceedings of the 2025 CHI conference on human factors in computing systems, Yokohama, Japan, 26 April-1 May 2025, pp. 1–20. ACM.

60.

C-L

T-T

Lee

C-T

, et al. Novel AMI in Zigbee satellite network based on heterogeneous wireless sensor network for global machine-to-machine connectivity. Electronics 2024; 13: 1421.

61.

Pliatsios

Kotis

Goumopoulos

. A systematic review on semantic interoperability in the IoE-enabled smart cities. Internet of Things 2023; 22: 100754.

62.

Bhattacharya

Peddi

Ponaganti

, et al. A survey on various security protocols of edge computing. J Supercomput 2024; 81: 310.

63.

Nakagawa

Antonino

Schnicke

, et al. Continuous systems and software engineering for industry 4.0: a disruptive view. Inf Software Technol 2021; 135: 106562.

64.

Mazlan

Mohd Daud

Mohd Sam

, et al. Scalability challenges in healthcare blockchain system—a systematic review. IEEE Access 2020; 8: 23663–23673.

65.

Hoek

Brouwer

JLP

Voors

, et al. Smart devices to measure and monitor QT intervals. Front Cardiovasc Med 2023; 10: 1172666.

66.

Petsani

Kehayia

Konstantinidis

, et al. Leveraging Pervasive Technologies for monitoring patients during hospitalization: a pilot study presenting the perspectives of the treating clinicians. Proceedings of the 17th international conference on PErvasive technologies related to assistive environments, 26-28 June. Crete, Greece: ACM, 2024, pp. 530–535.

67.

Rajendran

Porwal

Anjali

, et al. Portable IoT devices in healthcare for health monitoring and diagnostics. In: Murthy

Zurek‐Mortka

Pillai

, et al (eds) Internet of Things in Bioelectronics. Wiley, 2024, pp. 263–296.

68.

Tramontano

Tamburis

Cioce

, et al. Heart rate estimation from ballistocardiogram signals processing via low-cost telemedicine architectures: a comparative performance evaluation. Front Digit Health 2023; 5: 1222898.

69.

Fuller

HJA

Maddox

Adams

, et al. Considering team dynamics during human factors work in healthcare. Proceedings of the International Symposium on Human Factors and Ergonomics in Health Care 2023; 12: 156–159.

70.

Motta

Aracri

Ferretti

, et al. A framework for FAIR robotic datasets. Sci Data 2023; 10: 620.

71.

Benis

Grosjean

Billey

, et al. Medical informatics and digital health multilingual ontology (MIMO): a tool to improve international collaborations. Int J Med Inform 2022; 167: 104860.

72.

Koutalieris

Symeonidis

Kapsomenaki

, et al. Enhancing urban environmental sustainability through unified stakeholders needs co-creation process (AENEA). In: 2023 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE), Milano, Italy, 25–27 October 2023, pp. 899–904. IEEE.

73.

Tramontano

Scippacercola

Magliulo

. Estimating quality of life variation using BLE wearable devices on android based architecture. Int J Radiat Oncol Biol Phys 2018; 102: e286.

74.

Liu

Zhang

Zhou

. A comprehensive study of Bluetooth low energy. J Phys: Conf Ser 2021; 2093: 012021.

75.

Singh

Kumar

(eds) Security, privacy, and trust in wbans and e healthcare. CRC Press, 2024. DOI: 10.1201/9781032635101.

76.

HIPAA and GDPR . compliance in IoT healthcare systems. Communications in Computer and Information Science. Switzerland: Springer Nature, 2023, pp. 198–209.

77.

Koulouras

Katsoulis

Zantalis

. Evolution of Bluetooth technology: BLE in the IoT ecosystem. Sensors 2025; 25: 996.

78.

Shaikh

Jammal

. Survey of fault management techniques for edge-enabled distributed metaverse applications. Comput Netw 2024; 254: 110803.

79.

Ramakrishna

Ali Shaik

. A comprehensive analysis of cryptographic algorithms: evaluating security, efficiency, and future challenges. IEEE Access 2025; 13: 11576–11593.

80.

OneAquaHealth website. ONEAQUAHEALTH. https://www.oneaquahealth.eu/ (accessed 9 May 2024).

81.

Blobel

. Analysis, design and implementation of secure and interoperable distributed health information systems. Stud Health Technol Inf 2002; 89: 1–329.

82.

Soft systems methodology . Systems Approaches to Making Change: A Practical Guide. London: Springer, 2020, pp. 201–253.

83.

Blobel

Oemig

Ruotsalainen

, et al. The representational challenge of integration and interoperability in transformed health ecosystems. J Phys Math 2024; 15: 4.

84.

Kim

Choi

Lim

, et al. Business process-organizational structure (BP-OS) performance measurement model and problem-solving guidelines for efficient organizational management in an ontact work environment. Sustainability 2022; 14: 14574.

85.

Datta

Thomas

. The cube data model: a conceptual model and algebra for on-line analytical processing in data warehouses. Decis Support Syst 1999; 27: 289–301.

86.

Zhang

Jiang

. Response efficiency optimisation of data cube online analysis for network user’s behaviour. Int J Autonom Adapt Commun Syst 2023; 16: 270–283.

87.

Liyanage

Luzi

De Lusignan

Pecoraro

McNulty

Tamburis

Krause

Rigby

Blair

. Accessible Modelling of Complexity in Health and associated data flows: asthma as an exemplar. Journal of Innovation in Health Informatics. 2016;23(1):476-484.

88.

Garikapati

Turnbull

Bennett

, et al. The role of contemporary wearable and handheld devices in the diagnosis and management of cardiac arrhythmias. Heart Lung Circ 2022; 31: 1432–1449.

89.

Communications in Computer and Information Science. Standardization in cloud computing: unlocking the potential of a fragmented industry. Switzerland: Springer Nature, 2023, pp. 3–12.

90.

Kabir

Ray

. Virtual prototyping for modern internet-of-things applications: a survey. IEEE Access 2023; 11: 31384–31398.

91.

Dar

Olapade

Ottun

, et al. TOAD: profiling and evaluating 3D printed IoT rapid prototype designs. ACM Trans Internet Things 2025; 6: 1–31.

92.

Bourreau

Kraemer

. Interoperability in digital markets. SSRN Journal. DOI: 10.2139/ssrn.4181838.

93.

IFIP Advances in Information and Communication Technology . Design and development of a marketplace-based collaborative ecosystem for software integration and distribution within manufacturing. Switzerland: Springer Nature, 2024, pp. 80–95.

94.

Zhou

Park

Longardner

, et al. Clinical validation of a wearable ultrasound sensor of blood pressure. Nat Biomed Eng 2024; 9: 865–881.

A cloud–edge reference architecture for intertwining health digital domains

Abstract

Keywords

Introduction

Research background

LinkAll architecture design

Case studies

Hardware and software deployment summary

Remote monitoring system for urban greenery

Elderly individuals home care

Discussion

System performance evaluation

Architectural flexibility and scalability

LinkAll positioning in remote monitoring scenario

Limitations and future research directions

Conclusions

Footnotes

ORCID iDs

Ethical considerations

Consent to participate

Author contributions

Funding

Declaration of conflicting interests

Data Availability Statement

Conference submission

Appendix

References