The Cost Leverage of Early Decisions
There is a well-established principle in engineering project management: decisions made early in the design process are inexpensive to reverse; decisions made late are expensive. The conceptual design phase — the period from requirements definition to the selection of a basic design approach — typically consumes 5–10% of a project’s total effort but locks in 60–80% of the total product cost and most of the performance characteristics.
This asymmetry means that engineering time invested in the conceptual phase pays higher returns than the same time invested in detail design. A wrong material selection caught at concept costs a conversation; the same mistake caught after tooling has been ordered costs weeks of rework and potentially tens of thousands of dollars in wasted tooling.
This article describes the decisions that have the highest downstream impact and therefore must be resolved — with documented rationale — before the project moves into detail design.
Requirements: The Foundation Everything Builds On
The most expensive source of late-stage changes is requirements misunderstood or undiscovered at project start. Before any design work begins, every key requirement should be documented, categorized, and — critically — verified with the customer or stakeholder who owns it.
Requirements fall into three categories:
- Functional requirements: what the product must do (maximum force, operating speed, temperature range, fluid compatibility, cycle life)
- Constraint requirements: what the product must not violate (envelope dimensions, weight limit, interface compatibility, regulatory standards)
- Preference requirements: what the customer would like but which can be traded off against other requirements (cost, appearance, specific features)
The distinction between constraint and preference is critical — treating a preference as a constraint over-constrains the design space and drives unnecessary cost. Treating a constraint as a preference creates a compliance problem.
Architecture Decisions: High Impact, Hard to Reverse
The architecture of a mechanical product — how the primary load paths are structured, how the major assemblies are divided, how key functions are allocated to components — is the fundamental decision that constrains all subsequent detailed design. Architecture decisions include:
- Fabricated vs. machined from solid vs. cast structure
- Integral vs. modular configuration (one complex part vs. several simpler parts assembled)
- Location of the primary datum surfaces and how the product is fixtured for machining and assembly
- How adjustment and tolerance compensation are built into the assembly
- Where the wear items are, and how they are replaced in service
These decisions are difficult to change after detail design because they typically affect every drawing in the product. Changing from a fabricated frame to a cast frame at 70% detail design completion means discarding most of the completed drawings.
Material Selection at the Concept Phase
Material selection at concept does not require specifying a precise alloy and temper — it requires confirming the material family (steel, aluminum, polymer, composite) and the approximate property range needed. This is enough to commit to a manufacturing process and estimate cost.
The most common material selection mistake at the concept phase is defaulting to the material used in the previous similar design without evaluating whether it is still optimal for the new requirements. Changes in service environment, cost targets, weight targets, or production volume may make a different material more appropriate.
Material selection should be evaluated against: strength-to-weight ratio (if weight is constrained), machinability or formability for the intended manufacturing process, corrosion resistance in the service environment, cost at the required production volume, and supply chain risk.
Manufacturing Process Commitment
Manufacturing process selection at the concept phase determines the range of achievable geometries, the typical tolerance capability, the unit cost structure (high tooling cost with low variable cost for casting/injection molding vs. low setup cost with high variable cost for machined parts), and the supply chain required. Getting this right early prevents the situation where a design is completed to a casting-appropriate geometry and then machined (expensively) because the volume does not justify tooling.
| Decision | If Wrong at Concept | Cost to Fix at Detail Design | Cost to Fix After Tooling |
|---|---|---|---|
| Material family | All drawings specify wrong material | Medium — redraw and recalculate | High — possible retooling |
| Architecture (modular vs. integral) | Entire part structure wrong | High — most drawings must be redone | Very high — scrap tooling |
| Manufacturing process | Geometry may be non-manufacturable | High — feature redesign required | Very high — tooling scrapped |
| Interface/envelope dimensions | Product does not fit customer system | High — major redesign | Very high — possible product rejection |
| Service access and maintainability | Cannot be serviced in field | Medium — add access features | High — field modification campaign |
Design Reviews at the Concept Stage
A concept design review should be held before any detail design begins. Its purpose is not to approve aesthetics — it is to verify that requirements are correctly captured, that the chosen architecture satisfies all constraints, and that no fundamental technical risks remain unaddressed. The review should include manufacturing and service perspective, not just design engineers.
Document the review decisions, including alternatives considered and rejected with their rationale. This record is valuable when future engineers or customers ask why the design is the way it is — and someone always asks.
FAQ
Q: How much detail should a concept design have before moving to detail design?
Enough to confirm that the architecture is feasible, that the envelope constraints are satisfied, that the primary load path works structurally (verified by hand calculation or simple analysis), and that the manufacturing process can produce the key features. The concept does not need detailed dimensions, tolerances, or surface finish callouts — those belong in detail design. The transition criterion is: all major risk areas have been identified and the top two or three have been resolved or have a clear resolution path.
Q: How do I handle a customer who keeps adding requirements after concept design is approved?
Scope change after concept approval must go through a formal change process with explicit impact assessment on schedule, cost, and technical risk. Document this process clearly in the project agreement from the start. Requirements added after architecture commitment often force architectural changes — the most expensive category of change. The customer who adds requirements late must understand and accept the cost consequence, not receive the change for free through your engineering team absorbing overtime.
Q: How do I evaluate competing concept architectures systematically?
Use a weighted decision matrix: list the evaluation criteria (cost, weight, serviceability, manufacturability, risk, schedule), assign weights based on project priorities, score each concept against each criterion, and compute the weighted total. The process forces explicit discussion about which criteria matter most and prevents decisions by preference or by whoever speaks loudest in the review meeting. Document the matrix with the decision record — it provides the rationale for the chosen direction when it is questioned later.
コメント