Lean design programs cut waste that is hard to see.
Lean manufacturing gets lots of press and for good reason. It drives waste out of manufacturing facilities and makes them more profitable. It's relatively easy to see waste in manufacturing because you can touch it. But it's not so obvious in engineering departments. Still, lean-manufacturing principles apply there as well. In fact, engineering departments are good places to start lean-manufacturing programs.
The essence of lean, as it's often called, concentrates on removing waste while improving operations. Six Sigma principles, frequently mentioned in the same breath with lean ideas, are closely associated and focus on reducing defects, a form of waste. Together they form a notion of Lean Product Development/Lean Design, a method of removing waste from products before they get to the manufacturing floor.
Starting a lean-engineering program is beyond the scope of this article. However, one of its critical elements will be engineering software that eliminates the time lost to data translations, provides a database for reusing company ideas and best practices, and includes simulation and analysis software that eliminates prototypes.
LOOKING FOR WASTE
To find the waste in the engineer department, it's useful to first examine where waste comes from on the shop floor where it's easy to spot, and then work back to the design department. Experts say there are seven categories of manufacturing waste. They are listed in the accompanying table "Seven deadly wastes." Toyota often gets credit for compiling the list.
To illustrate how costs are unnecessarily added to designs, consider the production of railway rolling stock. An A.T.
Kearney study from 2003, The Line on Design, How to Reduce Material Cost by Eliminating Design Waste, found a total of 58% of the cost for internal production and material were wasted. More costs came from failing to use deign experience, overdesign, and making designs difficult to manufacture. The table "Waste in action" summarizes the finding and tells a sad tale of lazy management. The good news is that the total cost of the vehicle could be reduced by 30% over two years by simply attacking these areas.
If your company struggles with the following issues, it might benefit from improving its software and how it's used. Look for:
- Process delays and time lost looking for information, waiting for test results, and feedback.
- Unnecessary documents and physical prototypes.
- Designs never used, completed, or delivered.
- Poor designs that generate warranty issues.
- Under use of design knowledge, as in costly parts.
- Late identification of manufacturing errors.
SUPPORTING LEAN WITH SOFTWARE
In a nutshell, lean should help companies reduce delays, maximize design reuse, reduce defects, and improve process efficiency. One way to tie these ideas together would be with a design or modeling system that eliminates recreating data. Simulation in sufficient detail could predict performance without building physical prototypes. And a capable data-management system would track the mountains of information generated in modern projects. These capabilities support lean initiatives by eliminating the waste of delays, errors, and data loss through development stages, with improvements in overall process efficiency and cycle time.
The data-management system mentioned would be a structured repository for all product and process data required for development. This includes data created in computer-aided design, engineering, and manufacturing, along with information needed for development. Managed environments let companies capture workflows that route information to teams when needed, eliminating time wasted looking for, waiting for, and recreating product data.
A few other key software capabilities to complement lean engineering would include:
Associativity. In engineering software, it can eliminate the waste of missed updates.
Associativity often links CAD and CAM systems so that when a model updates, its NC work also updates. But when applied company wide, an updated model would kickoff updates everyplace that uses or mentions the model, such as manuals, drawings, and NC toolpaths.
Knowledge-driven automation. It lets manufacturers capture and reuse knowledge to automate development tasks. Knowledge-driven automation works several ways. It captures knowledge and best practices for specialized tasks such as structural analysis and mold and tooling designs.
Knowledge-driven automation programs would apply the experience of experts to frequent engineering tasks, such as stress analysis. In knowledge-driven automation, a rules-evaluation engine lets companies drive designs with external requirements and knowledge databases, those specific to the company and its products.
Simulation, validation, and optimization. Simulation lets engineers see how products carry loads, where designs are weakest, or how they react to heat, among other things. Comprehensive simulation predicts performance early in development, when a design is easiest and least expensive to alter. Structural and motion-analysis tools let users simulate early without requiring engineering analysis specialists.
Validation implies, for example, finding interference fits to eliminate manufacturing defects. And optimizations can be applied several ways. For instance, weight optimization takes unneeded material out of assemblies, and cost optimization, such as design for manufacturing, cuts the number of parts needed in an assembly.
System-based modeling reduces waste several ways. Take a jet engine, for example. High-level system modeling might run through a set of inputs and design-rules to define a framework, sheet metal, and volumes for subsystems such as a turbine section and fuel systems. Product or subsystem templates would then build on the outputs of the system model to detail the turbine section and fuel system.
System-design software lets companies standardize design practices and rapidly create product variants, reusing knowledge, and eliminating engineeringrework. Systems modeling also reduces engineering errors and defects among product options, variants, and derivative platforms by preserving highlevel design parameters and interfaces between systems and components.