Back to Articles
Design for Manufacturing
June 21, 2026

The 3-Question Rule That Can Eliminate 50% of Your Assembly Costs

NTE Lab

Hardware Engineering & DFMA

Most engineering teams unknowingly bleed profit on the assembly floor because of "part creep." Every bracket, screw, and cover added to a CAD model represents a new PO, a new supplier, a new tolerance stackup, and continuous labor overhead.

Before optimizing how a product is put together, you must ask a brutally simple question: Should this part even exist?

In the Boothroyd-Dewhurst Design for Assembly (DFA) methodology, the ultimate metric is Assembly Efficiency (E_ma), which mathematically compares your actual design against an idealized baseline.

The DFA Index Equation:

E_ma = (N_min · t_a) / t_ma

Where:

  • E_ma = Assembly Efficiency
  • N_min = The theoretical minimum number of parts
  • t_a = Basic ideal assembly time for one part (industry standard is generally 3.0 seconds)
  • t_ma = Total actual estimated assembly time for the product

💡 The Example

You design a module assembly with 12 distinct components (base, cover, PCB, 4 screws, 4 washers, 1 rubber seal). Your manufacturing team estimates a total assembly time (t_ma) of 85 seconds.

To find your true N_min, you must evaluate every single component against the DFA Minimum Part Criteria. Ask these three questions:

  1. Kinematics: Does the part fundamentally move relative to the parts already assembled?
  2. Material: Must the part be made of a fundamentally different material? (e.g., a conductive copper trace vs. an insulating plastic housing).
  3. Accessibility: Must the part be separate to allow for the assembly or disassembly of other necessary parts?

If the answer is NO to all three questions, the part is theoretically redundant and should be combined.

In our example, the screws and washers do not move, do not need to be a different material for functional isolation, and don't grant necessary access. They fail the test. Your true N_min is actually just 4 parts (Base, Cover, PCB, Seal).

Let's look at the true efficiency:

E_ma = (4 · 3.0) / 85 = 14.1%

The Result: You are operating at an abysmal 14% assembly efficiency. You are paying for 8 redundant parts and the massive labor overhead to orient and drive them.

🛠️ The Solution

Stop optimizing the installation of bad parts. Eliminate them entirely.

  1. Consolidate Geometries: Merge redundant parts into a single, complex injection-molded component.
  2. Eliminate Fasteners: Replace the 8 screws and washers with cantilever snap-fits molded directly into the rigid housing to secure the PCB and cover.
  3. Utilize Multi-Shot Molding: The rubber seal passes the material question, but it doesn't need to be a separate assembly step. Use two-shot (2K) molding to overmold the elastomer directly onto the plastic base, dropping N_min to 3.

Great engineers don't just design parts; they design the elimination of parts.

Have a product challenge like this?

We help hardware teams solve complex engineering, DFM, and production challenges every day.

Discuss Your Project