The concept of vertical integration falls under system integration. Through the process of putting together and connecting subsystems, the result is to create a fully functional unit where each subsystem benefits from the other. Teams use the term ‘silos’ to describe vertical integration, where each layer in the system works in an upward motion. The system is tightly integrated to ensure it serves a specific business function.
Only a few vendors, partners and developers make use of the most straightforward and quicker method for vertical integration. However, as the process evolves, this expands as processes change and newer functions get added to meet business requirements. An example of vertical integration is radiology. A Radiology Information System (RIS) looks at results as a Picture Archive and Communication System (PACS) and preserves these images. The data is held in place, and no other silos are coordinated.
The advantages of having vertically integrated solutions are:
- The supply chain coordination is effective
- All vendor communication is streamlined
- Improved competition in the market
- Users have more control over processes
- Reduction of operating costs
Horizontal integration, otherwise known as enterprise service bus (ESB), is a specially designed subsystem that the program allocates to communicate with other subsystems. It is a pattern where centralised software components complete integrations to backend systems. Included in these systems are translations of data models, routing and requests. Once this gets done, the integrations and translations are available as service interfaces which can be reused by any new application.
The ESB pattern has a specially designed integration runtime to ensure productivity.
Service-oriented architecture (SOA) relies on ESB as it is a crucial element in making software components reusable via service interfaces. The advantage of these interfaces is the fact that they use standard communication values to quickly and effectively incorporate new applications without having to run through a deep integration performance.
When running an SOA, it is essential to understand that each service carried out by SOA embodies specific code and data integrations to carry out discrete business functions, such as checking a client’s credit details or inspecting a loan payment.
The health care centre also makes use of ESB to assist in simplifying integrated medical systems. ESB can effectively work with customer systems through customised development.
Advantages of using ESB solutions include:
- Increased operational efficiency
- Improved business continuity
- Adaptable communication capabilities
- Increased scalability for large workloads
- Expansion of new technology
Before diving into what star integration is, it is crucial to explain point-to-point integrations. Linking one system to another to make them function together is basically what point-to-point integration is all about. Although it lacks complexity due to only requiring two components to perform one single function, they are instrumental in understanding star integrations.
Star integrations (also known as the spaghetti method) is a process of integration where tools are interconnected.
As the number of integrated systems increase, so does the number of integrations. Although most believe it to be a simple and highly flexible process, if the number of integrated systems goes over ten, the process of integration is very demanding (especially when working with heterogeneous and proprietary interfaces).
Companies would make use of star integration for the following reasons:
- Implementation is fast and straightforward
- Distribution and deployment are flexible
Common data format
Data is a crucial element in an organisation. Data management aims at effectively getting data to data consumers (human or applications).
Users can manage data interfaces through the use of data hubs. An enterprise data integration strategy, where hubs of data simplify the management of data flow between applications, is beneficial. Potential interfaces between applications result in an exponential function of the number of applications. So, organisations that have up to one thousand applications may have as many as half a million interfaces (this is assuming all applications are communicating with one another).
Through the use of hubs of data, users can reduce the number of interfaces resulting in a linear function of the number of applications.
Master data management hubs create a central place for all applications that require master data. Going to a data warehouse where all data gets stored in one place is an example of this kind of data management.
Data hubs assist in simplifying complex data interfaces. An example where data hubs help a lot is with formatted data. An updated application gets placed into a standard data format and only has to get configured into one particular format instead of many different types.