This paper discusses findings and challenges faced when implementing system modelling as a part of MBSE in a naval ship design process. The design, production, and life-cycle support of complex military vessels is accompanied by large sets of requirements. These sets are created, maintained, and processed by various stakeholders and face an increase in interrelations due to extended automation and electrification of ships. These challenges are addressed by advancing the design process to a model-based system engineering (MBSE) process. The current V-model based design process serves as the framework within which MBSE is pursued.
MBSE greatly improves traceability of requirements, consistency between system solutions and requirements, and product verification. Concretely, a MBSE implementation into the current design practice based on the Arcadia methodology and the Capella tooling is pursued. The main advantages are a well published methodology, wider application within the supply chain, and a community supported, open-source toolset.
The initial results include a successful integration into the existing design process. Hereby both the connection with the design disciplines such mechanical design, automation, and electrical design, and connection with the design aspect analysis such as vulnerability, safety, and ILS has been established. Within this transition the usual challenges were faced, including but not limited to setting up a new team, establishing the work processes, and the culture change that comes along with MBSE. These were addressed by training, information, and a robust work process.
However various challenges remain, especially with the support functionality that results from selecting specific physical solutions. The developed work process starts with evaluating the desired missions that the design should be able to execute. For each of these missions the required capabilities are modelled. Subsequently, for all capabilities a functional chain is modelled which links all required functions and their dependencies. However not all capabilities are directly linked to the missions as some capabilities are required to support other capabilities and often these support capabilities are only identified when specific physical solutions are chosen. Therefore, to develop a consistent set of capabilities, the design needs to be developed simultaneously at both capabilities level and the physical solution. This simultaneous development causes an iterative process which is insufficiently supported in the tools and a challenging control process as the design develops in a non-linear fashion. Nevertheless, a method was found to circumvent some of these challenges by pre-allocating certain support capabilities and their functionality in a logical distribution network. This allows for a more gradual development and thereby overcomes aforementioned challenges. But it predefines part of the physical solution already in the capability design. Also, it doesn’t prevent the iterative process, because the pre-allocation is as good as the designer’s foresight into the design development.
Concluding, a successful initial implementation of system modelling activities within the context of the current design process was performed. Additionally, a method was developed to deal with the interrelation between support capabilities and physical system solutions. The paper concludes with a forecast on future developments in this application.
