The Solution: Implementing a Digital Twin with Simio
Selecting the Right Technology Partner
When evaluating potential solutions, Westinghouse selected Simio for its flexibility and adaptability. According to Brad Parker:
“The flexibility, you know, the ability to add, take away… other models that I looked at seemed to be more the structure of it in terms… it was almost like hard coded, which made amendment difficult. [With Simio] it’s okay, we can be flexible enough to make it fit your process. Not your process has to change to reflect our way of doing it.”
This flexibility was crucial for handling the complexity across five sites with a vertical supply chain. Westinghouse partnered with Mosimtec, a consulting services company with over 14 years of experience in applying modeling and simulation across various industries.
Implementation Methodology
Mosimtec followed a proven process for deploying simulation models and implementing digital twins at Westinghouse sites:
- Functional Requirements Specification: Developing a detailed specification document outlining the system to be modeled, the modeling approach, and required inputs/outputs
- Phased Model Development: Building the model iteratively, enabling Westinghouse to see working interim versions early in the process
- Verification: Testing the model to ensure it met specifications and performed as intended
- Data Collection and ETL: Defining data requirements before development began, enabling parallel workflows
- Validation: Ensuring the model accurately reflected real-world system behavior
- Analysis: Gaining insights into system performance and behavior
- Integration and Change Management: Applying the model effectively to drive business improvements
- Ongoing Analysis and Support: Continuing to use the model for strategic, tactical, and operational decision-making
Digital Twin Architecture and Data Flow
The implementation integrated data from multiple sources to support capacity planning and scheduling:
- SAP: Provided information about resources, materials flow, bills of materials, routing orders, and inventory
- IMS: Contributed real-time data including equipment status, work in progress, and current tasks
- SAP Ariba: Handled data related to replenishment parameters, lead times, and reorder quantities
- Excel: Included labor qualifications, process data, and staff availability
An ETL (Extract, Transform, Load) functionality moved data from these sources into Simio, which processed it to produce capacity-feasible planning and scheduling. For rapid implementation, Westinghouse initially used an Excel Power Query-based approach as an interim solution.
Implementation Process: Building the Digital Twin
System Understanding and Definition
The implementation began with developing a detailed functional design specification. As Mosimtec explained, “If you can’t clearly describe a system, you can’t model it effectively.” This document outlined both the real-world system and the intended model using diagrams and narrative to convey the necessary level of detail.
The specification was reviewed and approved by Westinghouse before model development began but remained a living document that evolved alongside the model. Westinghouse found this document especially valuable as it provided a single, comprehensive reference capturing the entire process from start to finish.
Data Preparation and Integration
To build a true digital twin that extracted live data and ran scheduling plans, it was crucial to ensure the right data came from the right sources at the right time. Mosimtec developed ETL solutions to clean and integrate raw data, an iterative process involving onsite stakeholders with many adjustments made along the way.
Throughout model development, extensive use of error logs provided feedback to Westinghouse, helping them clean and standardize their data. The models created were fully data-driven, relying on real-time dynamic data inputs. This approach provided users with the flexibility to quickly refresh and generate new plans spanning multiple years.
Key Dashboards and Outputs
After each simulation run, users could access several key dashboards:
- Production Performance: A comprehensive overview comparing results from the generated plan to actual operational or contractual targets
- Material Flow: Visualization of material movement throughout the production process, ensuring it matched the bill of materials
- Throughput Analysis: Detailed throughput data by area, providing a clear view of how materials and resources were performing across different parts of the facility
- Resource Utilization: Statistics offering insights into how effectively resources such as machines and labor were being used, key for identifying bottlenecks
- Resource Gantt Chart: One of the most widely used dashboards at Westinghouse, providing a high-level view of how sales orders moved through various manufacturing processes
Results: Transforming Planning and Decision-Making
Dramatic Reduction in Planning Time
The implementation of Simio’s digital twin manufacturing solution dramatically reduced the time required for planning and scenario analysis. As Brad Parker noted:
“With the implementation of Simio, what we’ve been able to do is make these decisions a lot quicker. The simulation model can run and we can change parameters, as you know, in Simio based off the tables, based off the demand data.”
What previously took up to two weeks could now be accomplished in just a few hours, enabling faster decision-making and more efficient planning processes.
Integrated Planning Across Sites
The digital twin provided an integrated schedule for all areas, eliminating gaps in analysis and improving visibility across the supply chain. This integration was particularly valuable for Westinghouse’s central planning function:
“Being able to see okay, what does this mean, what’s your schedule look like now. Oh, I can see this is going to be light here. Does that impact, what impact, scenario build. It has allowed us an integrated planning function which previously we struggled with.”
Enhanced Decision-Making Capabilities
The implementation vastly improved Westinghouse’s ability to:
- Run “what-if” scenarios quickly
- Respond faster to customer requests
- Handle unplanned outages more effectively
- Assess the impact of prioritizing one customer over another
- Forecast constraints and areas needing improvement
Data Quality Improvements
The implementation also highlighted data quality issues and inconsistencies across systems. As Brad Parker explained:
“You wonder why you struggle in planning when you’ve got four systems telling you a different number. It doesn’t work well.”
By bringing data together in the digital twin, Westinghouse could identify and address these inconsistencies, improving data quality and decision-making across the organization.
User Empowerment
Westinghouse planners can now update all tables and run plans independently, without requiring intervention from Mosimtec. They only contact the consultants when results look strange or when they need new dashboards or changes to existing ones. This self-sufficiency has empowered the planning team to take ownership of the digital twin and use it effectively for day-to-day planning.
