A product design project goes wrong at the edges: a load nobody wrote down, a budget nobody checked against the process, a mold nobody asked about. Each one is easier to fix on paper than after parts exist.
These product design project basics cover the order of work that keeps those gaps from forming: a problem statement, a requirements list, the computer-aided design (CAD) model that carries the requirements, prototypes that test them, and inspection that proves them. If the idea is still very early, start with our primer on product design basics and come back here when you are ready to pin it down.
Product Design Project Basics: Start With a Clear Problem Statement
Every project needs a clear, honest statement of the problem being solved, not the solution. Who runs into this problem, how often, what have they tried, and why did that fall short? A design built without those answers can be technically competent and still miss the people it was meant to serve.
Keep the statement to two or three sentences. If it runs to a page, split it, because each problem needs its own requirements. The brief that grows out of this question, and the stages that follow it, are laid out in our guide to the product design process.
What Are Requirements, and Why Do They Matter?
Requirements are the written list of what the product must do, must not do, and is allowed to do. Sorting them into three groups makes the list far more useful than a single undifferentiated wish list.
Functional requirements are the easiest to write, because they flow straight from the problem statement. The product must hold a specific load and fit a specific space, or connect to a specific existing component.
Performance requirements take more thought. They cover how well the product does its job: how fast, how accurately, under what conditions, and for how many cycles before it wears out. Writing them down forces a real decision about what “good enough” means for this product, rather than defaulting to the highest spec anyone can reach.
Constraint requirements need the most care. Size, weight, cost target, available materials, regulatory standards, and the manufacturing process are easy to leave loose early on, and a constraint that surfaces late means rework. Most people do not know how things are made and need guidance there, which is the reason to bring a fabricator in while this group is still being written. We would write it last, once the other two exist, because a constraint only means something against a known function and performance target.
Barcade’s Quick & Crash controller is a good example of a performance requirement. The original handles, thin-walled molded ABS (acrylonitrile butadiene styrene), were no longer available, and whatever replaced them had to take repeated slams into the holster and the occasional drop. That demand is why the cast replacements use a high-impact urethane resin instead of a copy of the original plastic. The target was thousands of additional shots, stated as design intent rather than a tested cycle count.
A useful test for any requirement is to ask what specifically goes wrong if it is not met. If the honest answer is nothing serious, it is a nice-to-have, and it should be labeled as one instead of being treated as a constraint. A feature nobody required still has to be designed, built, and checked, so a short list that stays honest beats a long one nobody checks.
Picture a bracket for an outdoor fixture. It has to support 15 pounds without deflecting more than a millimeter, and that is the functional requirement. Surviving 5,000 open-close cycles is the performance requirement, and a 40mm envelope at under two dollars in volume is the constraint. A proposed decorative chamfer can now be checked against something specific: does it affect any of those numbers? If not, it is a nice-to-have and can be judged on its own merits.
How Do CAD, Prototyping, and Manufacturing Method Fit In?
Each one tests a different requirement: CAD turns the written list into geometry, prototypes check that geometry against the real world, and the manufacturing method decides what the geometry is allowed to be. The CAD model is also the reference every later stage works from, so it should carry requirements as dimensions instead of notes.
A 40mm envelope that exists only in a document is easy to break without noticing. One built into the model as a dimension stays in front of whoever edits the file.
Prototypes reveal what a screen cannot. How an object feels in the hand, how it looks at actual scale, whether a mechanism works the way the model suggested: none of that is easy to judge until something physical exists. Each round needs a question it is meant to answer and a definition of a satisfactory answer. Our guide to structuring a prototyping process covers what each round of prototyping should test and when it is time to stop iterating.
The manufacturing method sets which geometries are possible and what each part costs. That is why it belongs in the constraint group and not in a decision made after the design is finished. Injection molding makes the point: uniform wall thickness helps prevent sink, warpage, and voids, because uneven thickness cools and shrinks unevenly, and any face that runs in the direction the mold opens needs a slight taper, called draft, so the part releases without dragging. Find these late and the design gets reworked.
Quality Verification Closes the Loop
The last stage is checking that manufactured parts match the design intent, and it is where the requirements list earns its keep. First article inspection is the formal check that a part taken from the first production run meets every requirement on the engineering drawing; it need not be the very first part made. A requirement written as a number becomes a measurement. One written as an adjective, “sturdy” or “tight fit,” gives the inspector nothing to measure against.
Scan-based inspection captures millions of points across the whole part. Comparing the scan with the CAD gives a color map of where the part is within tolerance and where it is not. Our metrology and inspection page explains how scan-based checks fit into a project.
If you have a requirements list, a rough sketch, or only a problem statement, send it over. Call 718-557-9578 or contact us, and we will go through it with you and say where the design stands before you commit to tooling.
Sources: Protolabs on draft in injection molding; DuPont’s general design principles for molding; NTS Unitek on AS9102 first article inspection; Creaform on 3D scanning for quality control.
