Externalizing Data Models for Domain-Specific Workflow Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Business Process Management (BPM) solutions are domain-independent and platform-dependent, limiting business experts' involvement in design and governance due to the technical nature of languages like BPMN, and requiring costly and error-prone data migration when switching platforms.
Innovation Solution
A domain-specific and platform-independent approach is implemented, using a Domain Specific Language (DSL) to generate a business object database as an API, allowing users to define data models simply and connect them to domain-specific activities, enabling seamless data reuse across processes and platforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If BPMN language is used for process modeling, then processes can be executed and managed, but business experts cannot effectively participate due to technical complexity
Solution Approach 1:
The system segments the complex BPMN modeling task into two distinct layers: a high-level domain-specific language (DSL) for business experts and a technical BPMN layer for execution. The DSL is broken down into domain-specific concepts (e.g., Patient, Diagnosis, Treatment) that business experts can understand, while the technical complexity of BPMN is isolated to the translation layer. This segmentation allows business experts to work with simple, intuitive domain concepts without needing to understand underlying BPMN technicalities.
Solution Approach 2:
The patent introduces an intermediary translation mechanism that automatically converts DSL process models into BPMN executable models. This intermediary layer acts as a mediator between the business expert's simple DSL notation and the technical BPMN execution engine. The translation process handles complex mappings, validations, and transformations automatically, shielding business experts from BPMN complexity while ensuring executable process models are generated.
2Reliability
If platform-specific BPM tools are used, then processes can be executed on specific platforms, but data migration becomes costly and error-prone when switching platforms
Solution Approach 1:
The system creates a universal DSL layer that serves multiple BPM platforms simultaneously. The domain-specific process models defined in DSL are platform-agnostic and can be translated to different BPMN implementations across various platforms. This universality allows the same business process definition to be deployed on different BPM platforms without requiring platform-specific modifications, thereby enabling easy migration and multi-platform execution with a single source of truth.
Solution Approach 2:
The patent employs a copying strategy where the canonical process model exists in the DSL format, and platform-specific BPMN models are generated as copies or transformations of this single source. Rather than maintaining separate platform-specific models, the system copies the DSL definition and automatically translates it to the target platform's BPMN format. This copying approach ensures consistency across platforms and simplifies migration, as changes are made to the single DSL source and automatically propagated to all platform representations.
3Measurement precision
If domain-specific concepts are integrated into process models, then business accuracy improves, but model complexity increases
Solution Approach 1:
The system applies local quality by introducing domain-specific concepts and attributes only where needed in the process model, rather than requiring comprehensive domain knowledge throughout the entire model. Business experts can locally enrich process steps with domain-specific information (e.g., adding diagnostic criteria to a treatment step, or patient-specific attributes to an appointment step) without affecting the overall model structure. This localized approach maintains business accuracy at relevant points while keeping the general process flow simple and manageable.
Data Source
AI summary
A computer-implemented method and system provide for externalizing data in process execution environments. The method includes extracting a data model from a domain specification. The data model includes a set of business objects, at least some of the business objects being linked to a domain-specific activity type by a business object relation. A database is generated from the data model which is accessible through an application program interface. Provision is made for a user to generate a domain-specific process model with links to business objects in the domain specification. The domain-specific process model is transformed to a domain-independent process model which has access to the database through the application program interface at runtime.