The HORSE framework features a number of components realising specific aspects of the production process or providing the safety of the humans and machines. It communicates with the other systems as demonstrated in the figure below.
Figure 23: Context of the HORSE system
The customisation of the system according to the needs of the specific deployment requires understanding of the functions of the available components and interfaces, including the taxonomy of the alerts. These are available in details in separate documents. Few non-binding guidelines:
Existing software systems could be integrated by:
- direct utilisation of the middleware and HEG (HORSE Exec Global) interfaces and databases, or
- by introducing an adapter (see HORSE-BOSCH interface) that exposes the system functions in the messaging middleware.
- Hardware equipment and other devices could be integrated through the appropriate device or agent interfaces. Some of the existing implementations (e.g. ROS bridge, industrial and device protocols) could be used or a completely new adapter.
For an integration to the messaging middleware the new component needs to:
- implement a local WebSocket client,
- know the topics of the expected messages and the messages the other party is expecting;
- know the intended recipients or send broadcasted messages;
- know the senders of the incoming messages (or filter all incoming messages on filter and other properties)
- know the access parameters of the local broker (IP and port);
- know the semantic of the payload
- It is expected that the existing DB structures should be customised to support the specific product and agent definitions.