3 questions

An area that receives less attention is the hardware and operating system infrastructure. While Chapter 5 discusses operating systems, servers, processors, and other physical infrastructure components, the IoT framework intentionally remains vendor neutral. Instead, it focuses on how data moves throughout the system and how the various layers interact to support interoperability and scalability.

Challenges of Managing IT Infrastructure

The IoT framework and Chapter 5 both identify several important infrastructure and management challenges associated with implementing modern information systems, while also proposing practical solutions. One significant challenge is integrating legacy systems. As explained in Chapter 5, organizations often continue using older systems because replacing them is costly, even though these systems frequently create integration problems. The IoT article reaches a similar conclusion, noting that older technologies and disconnected workflows make it harder for hospital systems to share data with each other.

The IoT framework addresses this challenge through the Gateway Layer, which functions as a bridge between medical devices and existing hospital applications. The gateway enables older systems to communicate effectively with Electronic Health Records (EHRs) and other clinical applications by aggregating data, normalizing different formats, and relying on standardized APIs (Yadegari-Asadi et al., 2022).

Scalability is another challenge, as Chapter 5 emphasizes that information systems must continue performing efficiently as demand increases. The proposed architecture addresses this through its Management Layer and Cloud Platform Layer, which use load balancing and dynamic resource allocation to automatically distribute computing resources during periods of increased demand, such as emergency room surges (Yadegari-Asadi et al., 2022).

Finally, both sources recognize that security and privacy are critical concerns when managing sensitive healthcare information. The IoT framework incorporates a dedicated Security and Privacy Layer aligned with GDPR requirements. It includes role-based access through OAuth 2.0, device authentication, encryption, and federated learning, which allows AI models to learn from data across multiple devices without exposing patients' personal health information. (Yadegari-Asadi et al., 2022). Alharbe and Almalki (2025) emphasize the protection of patient information as essential in IoT-enabled healthcare because large volumes of sensitive data are continuously collected and transmitted across connected devices. They highlight blockchain technologies as one approach to strengthening data integrity and privacy, which complements the proposed IoT framework's use of authentication, encryption, and role-based access controls.

Organizational Information Integration

The organizational information integration is the central objective of the proposed IoT architectural model. The class announcement defines organizational information integration as the seamless flow of information across an organization so that business processes operate together instead of as isolated information silos. The IoT article reflects this same objective by identifying fragmented infrastructure and siloed clinical workflows as major barriers to efficient healthcare delivery and proposing a unified architecture centered on the Clinical Information System (CIS).

The framework demonstrates this integration in several ways. First, the IoT Gateway Layer serves as middleware. As described in the class announcement, middleware enables incompatible systems to exchange information by translating data between applications. The gateway performs this role by collecting data from a variety of medical devices, filtering and standardizing the information, and converting it into a common format that can be shared across hospital systems.

Second, the IoT framework incorporates service-oriented architecture (SOA) through its Application Layer. SOA allows new applications to connect using reusable software services and standardized APIs instead of requiring hospitals to rebuild existing systems when new technology is introduced. This makes it easier to add tools like telemedicine platforms while keeping the existing infrastructure intact.

Finally, the IoT architectural framework extends integration beyond the hospital itself by connecting with regional Health Information Exchanges (HIEs). This allows providers to securely share patient records with specialists and other healthcare organizations. So information is not limited to a single hospital system. This helps ensure that everyone involved in a patient's care has access to the information they need while still protecting patient privacy and supporting interoperability.

Article 3: Master Data Management Case Study

The third case study addressed organizational integration through the lens of centralized and decentralized master data management (MDM) platforms. The authors noted the benefits and drawbacks of each type of system for a particular healthcare organization composed of various organizational units. MDM platforms are critical for managing data that is accessed by many subsets of an organization. Additionally, and fortunately for this course, MDM platforms highlight several important concepts of IS architecture, such as IT infrastructure, business processes, information system types, and organizational integration.

Various Components of IT Infrastructure

The Haug, Staskiewicz, and Hvam article provides a few examples of IT infrastructure highlighted in Chapter 5 of the Pearson reading. For instance, the authors note the existence of a broader Cloud ERP system; there is discussion of data storage and organizational data flow processes, and there is also mention of challenges with specific data types in the broader IS architecture. Additionally, there is excellent detail on the various information systems the organization uses and how they interact, namely product management, environmental product systems, spare parts databases, and a few others. The article provides exceptional detail, specifically in Table 3, on which organizational units have read/write privileges within each information system. This serves as an important foundation in understanding how the business integrates its systems and, perhaps more importantly, offers insight into broader business processes.

One area where the article falls disappointingly short is that it never really delves into much detail regarding the specifics of data management and storage, enterprise software applications, or data hardware. Each of these components is critical to data management and storage. Table 4 offers some detail on how the data quality MDM project aimed to handle data updates, transfers, and maintenance. For instance, when considering the issue of “incomplete item values,” a proposed concept was: “...when comparing data from two different IT sources, some values in less important attributes in product master data may be missing, which most likely happened when moving data from one system to another.”(Haug, A., Staskiewicz, A. M., & Hvam, L. (2023). While this highlights an issue with data quality writ large, it does not tailor the issue to the specific company. To better address this issue, it would be helpful to know 1.) what systems does the company use? 2.) If there are several, is there a combination of systems or platforms that tend to have compatibility issues? And 3.) Do all local organizational units tend to use the same systems? These details may seem mundane, but having the specificity of operating system, software, hardware specifications, etc. is critical to a thorough analysis of centralized and decentralized master data management systems.

Challenges of Managing IT Infrastructure in MDM

Given the main discussion in the Haug, Staskiewicz, and Hvam article concerns master data management (MDM), specifically whether decentralized or centralized offers a more effective solution, there are obvious challenges associated with scalability. If the firm in question opts to decentralize the management of the master data, there are issues as to whether access to that data and uniform maintenance of the data can be achieved in a larger organization. If each organizational unit is responsible for its own data, there is an obvious question of how reliable and consistent data is across the broader enterprise. Centralized management of the master data comes with its own challenges, however. While data consistency should not be as big an issue in a centralized platform, technological and hardware challenges become more prominent. If all data is managed at a central location, the organization must have the computing and information resources to process that data. Relatedly, there must be adequate software and business processes to ensure data from various IS systems is integrated throughout the infrastructure.

Related to scalability in this case study is also the question of cost. In a decentralized system, the cost of managing and maintaining master data may be less, but organizational units may be more likely to use external vendors for cloud storage or software as a service, each of which carries their own challenges. Separately, in a centralized MDM system, the total cost of ownership would likely be much higher than that of a decentralized system mainly due to the hardware, energy/utility costs, maintenance fees, and personnel.

One way to address these challenges is to ensure that business strategy and organizational goals are properly aligned with the investment in IT infrastructure and an organization’s IS architecture. Put more simply, if the business prioritizes agile decision-making because their competitive advantage is the ability to move quickly and adapt to competitors in the healthcare sector, then a decentralized platform may be more appropriate. In contrast, however, if the healthcare company is larger, more corporate-focused, and prioritizes internal IT innovation, then a centralized platform would likely be the most effective path.

This concept of aligning business goals with IS processes is not new; it is a quite common strategy of CIOs for businesses around the world. In other words, management and executives must familiarize themselves with the how data moves in an IS system, ensure data quality, account for specific business processes, and then use that information to determine what method of IS architecture works best for the organization. (Bergamo, Arriaga, 2011)

Organization Information Integration

Organizational integration is especially critical when it comes to master data, specifically the management, storage, and retrieval of such data. Master data is data that is used and accessed by many parts of the organization, such as customer information, supplier information, locations, and accounts. The article goes in depth into some issues that arise when the master data is of poor quality. For example, different master data tagging can lead to disruptions in the supply chain wherein one aspect of the organization has a different perspective and understanding of the state of input/output transactions than another. The same can be said for inventory, logistics, and product management. One practical example of mismanaged master data the article describes is if product weight data is tagged in one unit system in one country where the organization has an office and another unit system in a separate country where an office is located, the result can lead to confusion, wasted time, or worse, errors in production.

Quite interestingly, there were some significant challenges associated with centralizing the master data management platform in the referenced case study. (Haug, A., Staskiewicz, A. M., & Hvam, L. (2023). The first issue, a learning curve associated with navigating the system, was to be expected. Whereas business units previously used local systems, and therefore gained a sense of familiarity with such systems, the new platform inevitably led to confusion and questions as to where the data was physically and logically located. Second, and perhaps more surprising, was that the centralization of the master data increased lag time in locating missing data because business units now had to reach out to a centralized authority to help them find the data as opposed to reaching out to the decentralized units that had originally created the data. The authors of the study noted that the centralized authority GMDM often lacked the expertise that resided with decentralized units and therefore struggled to resolve data quality issues in a timely manner.

While these issues are certainly a risk of centralized management, the focus on “centralization” as the main culprit may be misplaced. If GMDM had expertise in-house, proper data infrastructure, integration, and business processes to oversee such components, many of the issues present in this study may have been absent, or at the very least, significantly mitigated. The takeaway from this case study is that master data management exists on somewhat of a continuum. The advantage of decentralized MDM is that problems with data quality can often be quickly addressed and expertise can be found locally. The drawbacks to decentralization are that integration of data across the organization is less likely to occur, and data quality could suffer. For centralization, data quality and integration can is a positive. Separately, centralization of MDM and broader integration can help foster organizational resiliency by allowing big data analytics to forecast economic, geopolitical, and broader organizational downturns (Cheng, 2025). Conversely, centralization can lead to a lack of expertise and data bottlenecks.

As with all of the aforementioned case studies, the key input with any effort to integrate IS infrastructure is the quality of the data. Again, consider the importance of integration within the supply chain system. As information flows increase between business units, the organization is able to more efficiently process customer requests, respond to changing customer preferences, and track business trends (Rajaguru, Matanda, 2009). Each part of the data flow is dependent on the other. This, however, also means that should the data be compromised, each layer of that system would also be compromised with it. Data quality and security of that data are so critical to ensure both survivability in the event of an outage but also for the purpose of maintaining data integrity.

Reference

Alharbe, N., & Almalki, M. (2025). IoTenabled healthcare transformation leveraging deep learning for advanced patient monitoring and diagnosis. Multimedia Tools and Applications, 84, p. 21331–21344. doi:https://doi.org/10.1007/s1104202419919w

Azubuike, C., Sule, A. K., Adepoju, P. A. A., Ikwuanusi, U. F., & Odionu, C. S. (2024). Integrating

SaaS products in higher education: Challenges and best practices in enterprise

architecture. International Journal of Research and Scientific Innovation, vol. 11(12), p. 948-957.