Apis system architecture
The Apis system architecture is a component based architecture. Apis systems and applications are glued together by assembling prefabricated components. All subsequent applications and systems presented on this site is built using the same component architecture.
Apis applications are software applications that delivers a specific set of functionalities on one computer. Any Apis application can be extended with new functionalities by adding new components. Components can be added later than initial setup, and often while the application is running.
Apis systems are distributed sets of Apis applications installed on various computers that interoperate to deliver the full set of functionalities. All Apis applications are highly distributable. There might be various reasons for distributing the application setup. Some applications is inherently distributed, like the Apis Distributed Historian. Uptime of critical functionality might be another reason for distributing functionalities that otherwise would run in one application or one machine.
Apis core system architecture diagram
The core system architecture diagram is set up to illustrate the various roles of the Apis components in a tiered architecture. The Data tier at the bottom involves components specializing in capturing or storing data. Apis Hive is the key component for data capture. It is an executable, often run as a windows service, that hosts various data components for communicating with other system over various protocols and technologies. All components plugged into Hive runs within the Hive executable, and thus have the most efficient access to data possible. The data in Hive is all current values, for instance the latest value of a sensor measurement or latest state of an alarm. The Logger component will stream the current values to the specialized time-series database, Apis Honeystore, where as the Alarm Server and the Tracker will store alarm states and process transactions to the Operational Data Store, ODS, a ISA-95 based fixed schema set in Microsoft SQL Server. On the data capture side, there are support for numerous standard protocols, such as OPC, WITS ML, SEMI-SECS/GEM/PV2, fieldbus protocols and DB/text file logs. There are also concepts for adopting new unknown/proprietary protocols efficiently.
The next tier in the architecture, is the Application Tier. The Apis Hive lies in the borderline between the data tier and the application tier, to assure efficient access to data from the application components within the same executable. This makes it possible to design applications, with real-time, deterministic processing requirements. Among the application components, we have:
- Apis ModFrame – a modeling framework for graphical design of mathematical dynamic models. Among applications designed with ModFrame are real-time models with estimates of non-observable states, so-called software sensors.
- Apis SoftPLC – a software based programmable logic controller with graphical programming in structured function charts, SFC, and function block diagrams, FBD, as defined in IEC 61131-3/PLCOpen.
- Apis Cameleon – a real-time semantic mapping component, that can map objects from a proprietary namespace structure to a domain specific namespace structure like ISA-95. The domain specific namespace, is then exposed through the server interfaces available in Hive, such as OPC UA.
All data captured and/or generated in Hive are exposed through the real-time Hive server interfaces, which are OPC DA, OPC AE, OPC HDA, OPC UA, SEMI SECS GEM, SEMI PV2 or the proprietary APIS Client Library. All data stored in the Honeystore time-series database, are exposed through the following server interfaces OPC HDA, OPC UA HA, ODBC, OLEDB or WCF. The data in the ODS can be extracted and loaded into an Analytical Data Store (ADS). The ADS is a SQL server based data mart, specialized for intuitive and efficient reporting and analyzing purposes. All general 3rd party Business Intelligence tools in the market will easily interface and consume the star schemas in the DB.