Evaluating Design For Manufacturing
You may already know a little about how we approach a design problem. Whether it be designing spreader bars and lifting beams or equipment for use in low temperatures, engineers at Sparta are always evaluating design and improving the process design for manufacturing (often referred to as “manufacturability“).
Anyone familiar with product development and engineering design will know that there are two absolute truths:
- First, there are almost unlimited ways to solve a problem (“millions of ways to skin the cat”) and that the process of arriving at the final design is a very iterative process.
- Second, when we hold something we often intuitively know when it is well designed (think of how holding your first iPhone made you feel).
What I hope to do is to outline a couple of easy steps that can be implemented on the fly during the design process to test the manufacturability of the product.
Evaluating Design on The Fly
After ten years of experience, and having floundered through my fair share of design problems, here is our quick process for evaluating a design:
1) Get rid of the design matrix
2) Understand the real problem
3) Count the number of parts
4) Record the amount of welding (in inches)
These steps are explained in detail below.
1. Throw the Design Matrix Out the Window
The first thing taught in many design classes is some form of design matrix that can be filled out in a table format that allows you to use weighted averages to compare required parameters vs time/money.
Stop using this immediately and throw it in the garbage!
I have never met a good designer that was effective at filling out tables and spreadsheets, nor do I think it helps the creative process to try such a blunt approach. Instead, use an approach that is intuitive, purposeful, and flows like the design process itself. Look at the product and make observations about how the pieces work together.
2. Understand The Real Problem
The first step to evaluating a design is to understand the real problem you are trying to solve. Interview the user/customer. What do they need? What is working and what isn’t? It takes a certain amount of skill to uncover the true problem. A customer may come to us with a list of parameters or specifications that need to be achieved and a price point. These are helpful in identifying the constraints of the project but not in determining the real problem. This information doesn’t consider what manufacturing design engineers are hired to do. We design solutions to manufacturing problems. We don’t design new products unless there is a problem or an inefficiency with the old product design.
During the first meeting with the customer, we try and guide the conversation towards discovering the data we need.
Watch out for value statements from the customer such as “We need Product X to be cheaper than competition A”, “We need to improve the safety of this device” or “we need to be able to do A, B, C faster and more efficient”
Ask lots of questions and gather data to determine the root problem worth solving.
3. Count the Number of Parts
This may seem intuitive at first, but when you are looking at a complex project you can quickly evaluate a design decision based on the number of parts required to solve the problem. I usually count the number of unique parts or the ones not used anywhere else in the design. The fewer unique parts the better. I love SolidWorks’ AssemblyXpert tool for this because in a split second it can output all the critical part counting metrics that are important to track.
In most cases and in most manufacturing facilities, having one part serve multiple functions or using fewer parts to finish your design will equate to a more efficient, cheaper, and better overall design.
4. Record the Amount of Welding
Here in Alberta, labor is still the most expensive portion of a manufacturing project. One of the quickest and easiest ways to predict labor requirements for a project is to count the amount of welding in inches. It depends a little on the complexity of the weld (for example, a pressure weld vs a sheet metal stitch), but if all things are equal, eliminating a weld will increase the efficiency of the design from a manufacturing perspective. If you have access to modern CNC cutting machines and metal bending equipment, removing welds from a project can be fairly easy to accomplish with a little thought. It also removes the possibility of human error when attempting to measure materials and welding two pieces together correctly.
Evaluating the Design Process For Manufacturing
We are constantly asking ourselves: Is this a good design for manufacturing?
For products like the free standing service rig or the new high tensile steel service rig mast, we use the quick evaluation method outlined above to evaluate the old product and determine if there is a more efficient way to design the pieces for manufacturing. For these two products in particular, advances in welding and material technologies enabled us to identify design solutions that would be more efficient. We determined moving forward with a more comprehensive design process and a re-engineering of the free standing service rig and service rig mast would improve the design for manufacturing. This allowed us to be recognized as industry leaders in service rig and workover rig designs.
Thanks for taking the time to read about our process. I find that when you explain our approach to people their first reaction is to just nod their head and say that its obvious. However, in my experience when you are mid-way through a design and you are trying to decide if path A or path B is better for manufacturing and better for the overall design of the product, it is easy to forget these simple tricks to help you make up your mind.