Engineering Design Transformation from fragmented groups to one operating model.
Building a global, collaborative design organization around common standards, shared workload, Siemens NX, and a controlled Teamcenter environment—while improving output, quality, and engineering-data discipline.
The transformation began after my June 2014 promotion to Director of Engineering and Design. Before the promotion, Trico President and CEO Lou Braga asked what my vision was for the design organization. I proposed a global group that could work together in a shared Teamcenter environment, move work across traditional business-unit boundaries, and demonstrate measurable improvement in efficiency and quality. I was given broad latitude to put that model into practice.
Make one design organization out of several separate groups
The design department had grown around customer and product business units rather than around a common operating model. Motor Engineering designers were primarily CATIA users. Aftermarket resources were still transitioning out of I-DEAS and into NX. Other OE system designers were largely NX-based, while the Texas resource used a mix of I-DEAS and AutoCAD. Offshore designers in India supported both CATIA and NX drawing work.
The largest problems were the number of CAD platforms, business-unit silos, inconsistent design and release practices, a stalled Teamcenter Unified Architecture rollout, inefficient use of offshore resources, and a large drawing backlog. The objective was not merely tool consolidation. It was to make the department function as one organization.
Break the silos, pool the capability
The first change was organizational. Designers were consolidated under a common supervisory structure. My design supervisor and I assigned work based on capability, workload, and relevant customer knowledge rather than treating each business unit as a closed resource pool.
That made underused capacity visible. If a motor designer did not have enough active development work, related motor ECR and release work elsewhere in the portfolio could be assigned instead. Department-level metrics were not a perfect measure of development engineering, but they provided enough visibility to manage work actively rather than allowing isolated pockets of unused capacity.
- Customer / business-unit silos
- Multiple CAD practices
- Local work queues
- Uneven utilization
- Common supervision
- Capability-based assignment
- Shared workload visibility
- Common standards and reviews
- Interchangeable resources where practical
- Collaborative design reviews
- Department-level workload control
- Measurable output
Turn good design practice into an auditable system
The organization developed a published design manual covering requirements by component, subassembly, and assembly type. It also captured practical engineering knowledge: recommended mold draft, component-specific design guidance, coordinate-system conventions, and other details that helped designers start from established product knowledge rather than rediscovering it.
A living drawing-design standards library maintained one approved reference drawing for each component or component type and each assembly type. The references standardized views, dimension placement, tolerances, and notes so Manufacturing and Quality knew where to find critical information and were less likely to miss a requirement.
The design-and-release procedure became a published, auditable process reviewed during internal and IATF audits. Designers completed and retained design checklists. Peer and supervisor reviews occurred during development and release. Teamcenter workflows enforced the required steps, and I approved every drawing.
A controlled drawing system, a manual CAD vault, and an abandoned TcUA migration
Teamcenter Enterprise 3.11 remained the controlled source for released 2D drawings, BOM records, temporary deviations, and approved ECRs. Physical control of released CAD was separate: Document Control managed directory check-in/check-out with paperwork. A designer requesting a change received a working copy; after approval and property updates, CAD was checked back into a protected directory. The link between released CAD and the controlled TCE drawing was procedural rather than a native PLM relationship.
TcUA 9 had been installed and an automated data load attempted, primarily for legacy I-DEAS content, but the migration was incomplete, contained errors, and had not been maintained. By the time the transformation began, it could not be trusted as current product data.
Legacy control
TCE 3.11
Released 2D drawings
BOM records
ECR and temporary-deviation approvals
Document Control directories
Manual CAD check-out / check-in
Protected released files
Target engineering environment
TcUA
Controlled CAD
Verified BOM structure
Drawing records
Release approval
Parallel ECR workflow
No reliable master existed, so migration became a configuration audit
The first implementation phase deliberately kept TCE operating while TcUA became the controlled CAD repository. Licensing economics made an immediate enterprise-wide cutover impractical, and the legacy information was not reliable enough to automate.
We created a master migration list with active products at the top and distributed migration work across the design organization. Migration was performed by assembly because the BOM relationships were essential to engineering-change impact analysis. Each migration reconciled actual product configuration, the TCE BOM, the controlled drawing, native CAD, the TcUA BOM, drawing record, and correct CAD revision.
Components and assemblies were migrated in their native CAD format along with an NX-translated part. Completed assemblies went through a TcUA release workflow and were double-checked before release.
Use live engineering work to pull active products into the new system
The modified ECR process became a second migration path. If an engineering change affected any component or assembly that had not yet been migrated, the affected CAD and product structure had to be migrated, reconciled, and released in TcUA before the ECR could be completed.
This combined planned migration by product priority with event-driven migration generated by current engineering work. It allowed the organization to improve the data continuously without a risky big-bang conversion.
Learn BMIDE and build the engineering controls we needed
Dedicated implementation resources were limited, so I worked directly in BMIDE rather than limiting my role to process definition. I created Trico-specific business objects, Trico Design Items, data types, custom fields, ECR and Temporary ECR objects, and many of the workflows used for release and approval.
I deployed BMIDE updates to the sandbox environment for development and testing. Internal IT handled production deployment while that organization existed; later production changes were supported through an external service arrangement. I remained hands-on with sandbox configuration and testing.
The engineering/design environment ultimately supported controlled CAD, drawing control, release approval, and a parallel engineering-change workflow in TcUA while TCE remained in service for broader users whose access was constrained by TcUA named-user licensing cost.
Replace nominally cheap capacity with productive internal capability
The India contractor model was not delivering the expected business benefit. Poor work quality, ineffective adoption of the new standards, supervision burden, and rework were outweighing the nominal hourly labor-rate advantage.
Six India contractors were replaced by three direct Trico designers on the Matamoros plant payroll. They worked from office space at the plant under the same design standards, Teamcenter process, workload controls, and KPI structure as the rest of the department. Day-to-day supervision came from Texas while I retained responsibility for the overall design organization and restructuring.
Standards became the foundation for smarter starter models
Common modeling discipline created benefits beyond drawing quality. Coordinate-system rules defined where the principal origins and alignment systems belonged for components and assemblies. New parts therefore entered assemblies predictably rather than requiring each designer to reconstruct placement logic.
That discipline eventually supported the smart-model initiative. Standard starter models could accept the principal kinematic points describing a wiper system and rapidly generate representative geometry good enough for customer technical presentations and fairly precise packaging evaluation.
Higher output with a smaller, more disciplined organization
The results came from the combination of organizational consolidation, common standards, verified PLM migration, workload visibility, and the redesigned resource model—not from any single software tool.
The drawing-backlog KPI was tracked by part number; charted drawings could therefore clear multiple backlog items with one released drawing. The organization nevertheless used the measure consistently as a practical workload and output indicator.
The transformation was an engineering operating-model change, not an IT project
The original vision was a global design organization that could work collaboratively in a shared environment, use resources flexibly, improve quality, and demonstrate better output. Achieving it required changes to organization design, technical standards, engineering governance, CAD and PLM practices, data quality, workload management, staffing strategy, and culture.
Teamcenter was an important enabling system, but the measurable improvement came from redesigning how the engineering organization worked.