HOW CUSTOM SPORTS EQUIPMENT IS DEVELOPED

2026-10-12 · for founders & owners · 14 min read · endothermal systems

Aknazar Arysbek, founder — biomedical engineer (MSc, Tufts University). LinkedIn ↗

Suppose you coach lacrosse. Practice ends with a hundred balls scattered across the field. Somebody has to pick them up. You have an idea for a machine that would do it. This note shows how sports equipment development takes that idea from a sketch to a product in a box, with the numbers worked out at every step.

Three things to settle first

Every project to build custom sports equipment starts with three things: intentions, constraints, and a system.

The intention is what you want the product to change. For our machine it is simple. The coach should stop spending the end of every practice bent over a field of balls.

The constraints are the limits the product must live within. Each one works like a fence around the design. Our machine must work without a motor. It must be run by one person. It must fit in a car.

The system is the method you repeat at every decision, and ours is the method of the ideal outcome. In it, you picture each possible solution working perfectly within the laws of physics, then judge it. If even the perfect version does not satisfy you, drop it; if it does, work out how to get there. This method throws out bad ideas while they are still cheap.

The six steps of sports product development

Every piece of hardware, from a helmet to a ball picker machine, passes through the same six steps. Each step answers one question.

We will follow one machine through all six: a ball collector for small balls. It has moving parts. It touches the ground. It rides home in a car. It also has to cost less than the alternative, which is a tired assistant coach with a bucket.

We will look at two sports, lacrosse and squash. Both use small balls. The differences between those balls decide much of the design.

Step 1: requirements

Before any drawing is made, you write down what the machine must do. This list is the requirements. Later decisions are checked against it.

A good requirement has a number in it. "Picks up balls quickly" is a wish. "Picks up 100 balls in 10 minutes or less" is a requirement.

Start with the ball.

The lacrosse figures come from the ball standard, NOCSAE ND049. It sets a weight of 5.0 to 5.25 oz and a circumference of 7.75 to 8.0 in, and the metric values are our conversion. The squash figures are the World Squash ball specification as published.

The lacrosse ball is about 1.6 times as wide as the squash ball. It is about six times as heavy. A machine built for both is harder to make than two machines built for one each.

For this example, we develop a lacrosse ball collector. Here is the list.

The numbers in R2, R3, R5, and R8 are assumptions for this example. In a real project they come from you, the person who stands on the field every day.

Now we put numbers on the requirements, and the list of wishes turns into an engineering problem you can solve with a pencil.

How heavy is a full load?

One hundred lacrosse balls at about 145 g each weigh about 15 kg. That is one giant watermelon. The frame, the wheels, and the handle must carry it. The operator must push it.

How much room do the balls take?

One lacrosse ball has a volume of about 135 cm³. Balls poured into a container leave gaps between them, and a random pile of spheres fills only about 64% of the space it occupies (random close packing). So 100 balls need about 21 liters of space. That is the size of a small backpack.

How fast must you walk?

Suppose the machine picks up every ball in a strip as wide as itself. Clearing a 30 m by 30 m area with a 0.6 m wide machine means walking 1,500 m. At an assumed steady walk of 1.2 m/s, that takes about 21 minutes. That fails R2. A 1.2 m wide machine halves the walk to about 10 minutes. But a 1.2 m machine must fold to meet R5.

So two requirements fight each other. Here there is a way out. Suppose the balls lie in clusters, as they might near the goal after a shooting drill. Then you walk from cluster to cluster. You do not sweep every square meter. R2 can hold, but only if its test spreads the balls the way they really lie. The test method in R2 must say how.

The rule: check your requirements against each other before you check them against the world.

Step 2: concepts

With the requirements set, the designer sketches every reasonable way to meet them. These sketches are the concepts. All of them go on the table before anyone picks a favorite.

Our collector has four concepts. Every one of them fails somewhere. The job is to choose the failure you can handle.

Now apply the method of the ideal outcome. Picture each concept working perfectly.

The perfect pickup tube still lifts one ball per stoop. One hundred stoops is the problem we started with. Drop it.

The perfect towed sweeper still needs a vehicle on the field. A team without one cannot use it. Drop it.

The perfect robot finds every ball. It also needs a battery, motors, a camera, and software that tells a ball from a dandelion. Even in its perfect form it fails the price limit in R8. Drop it.

The perfect push roller collects every ball in one pass with no motor at all. It passes. The verdict is concept B.

Notice why it won. Of the concepts that pass, B has the fewest parts. Each part you remove is one less to design, buy, or break.

Concept B is also on sale already, in a smaller size. The Kollectaball K-Field is a wire-cage roller made for lacrosse, baseball, field hockey, and cricket balls. It holds up to 25 balls and lists at $99.95. It proves the principle works on a lacrosse ball. It fails R3: a hundred balls means emptying it four times. The work that remains is a cage four times larger. Check the shelf before any project goes further.

Inside the push roller

The cage is a drum made of spring wires stretched between two end disks. As the drum rolls over a ball, two neighboring wires press down on it from above. The ball is wider than the gap between them. So the wires bend apart, and the ball pops through. Then the wires spring back. Now the ball is inside and cannot fall out.

Everything depends on the gap. Make it too wide, and balls fall back out. Make it too narrow, and the wires cannot bend far enough to let a ball in. A useful starting point is a gap of about 80% of the ball diameter. For a 64 mm lacrosse ball that is about 51 mm. For a 40 mm squash ball it is about 32 mm. Both numbers are assumptions to be tested in step 3.

Now the squash ball comes back. With a gap sized for the lacrosse ball, the squash ball falls straight through. With a gap sized for the squash ball, the lacrosse ball never gets in. One cage cannot serve both sports. A squash ball collector needs its own cage, or a frame that takes swappable cages.

Squash raises a second question. A court holds perhaps 30 balls inside four walls. Apply the ideal outcome again. A perfect machine saves a squash coach a minute or two. A basket may already do the job. The cheapest machine is the one you avoid building.

The rule: write down the smallest and the largest object the machine must handle before you draw anything.

Step 3: proof of concept

A proof of concept is a rough model built to answer one question. Does the central idea work?

For the push roller, the question is narrow. Do lacrosse balls enter the cage and stay inside when it rolls over grass? The handle, the paint, and the folding can wait.

The model can be a few spring wires bent around two plywood disks. Roll it over a patch of grass with ten balls on it. Count how many end up inside. Then change the gap by 2 mm and roll it again, across a range from 45 to 57 mm. Plot the count against the gap. The best gap is the one where every ball goes in and none come out.

Grass adds a second test. A lacrosse ball sits down in the grass. The cage must press the ball hard enough to push it through the wires before the ball simply rolls away. That push comes from the weight of the drum. Too light a drum skips over balls in long grass. Test on the longest grass the machine will meet.

This is the cheapest place in the project to be wrong. An idea that fails here costs some plywood, a coil of wire, and a few afternoons on the grass. The same idea failing after the molds are cut costs thousands.

Step 4: engineering prototype and testing

Once the idea works, the designer builds a prototype that looks and behaves like the real product. Every part is first drawn in a computer model. Each dimension, material, and fastener is listed.

The prototype has five main parts. Each one has its own engineering question.

Take the wires as an example. Every ball that enters bends two of them. Bend a paper clip back and forth, and it soon snaps. Metal tires under repeated bending. Engineers call this fatigue.

So count the bends. Assume 100 balls per practice, five practices a week, and a 30 week season. That makes 15,000 balls, and each ball bends two wires. Spring steel kept well below its limits handles far more bends than that. The wires also roll over wet grass. They must be stainless or coated against rust.

The prototype is then tested against the requirements, one line at a time.

Most prototypes fail at least one test the first time. That is what testing is for. Each failure is written down. The design is changed. Then the test runs again. Plan for two or three rounds.

Step 5: design for manufacturing

A prototype is built one at a time, with what you have. Factories build products by the hundred or the thousand. Design for manufacturing is the step where the design changes so a factory can make it quickly, cheaply, and the same way every time.

Here you meet the number that decides most hardware budgets: tooling. Tooling is the molds and fixtures a factory needs before it can make the first part. You pay for it before a single unit ships. A mold for one plastic part runs from about $1,500 for a simple aluminum tool to $100,000 for a large, complex one.

The cost of one unit is:

C is the cost of one unit. N is the number of units made. T is the tooling cost, paid once. M is material and parts per unit. L is assembly labor per unit.

Compare two ways to make the collector's end disks and frame parts. Method 1 prints them on a 3D printer and assembles by hand. Its tooling is the printer. Method 2 uses injection molded parts. All figures are assumptions for this example, not quotes.

At 100 units, each molded collector carries $300 of mold. At 5,000 units, the same mold adds $6. The two methods cost the same at about 516 units. That is the $29,400 difference in tooling divided by the $57 that molding saves on each unit.

So the way to build a product depends on how many you expect to sell. A league that needs 50 collectors should not buy a mold. At 5,000 units, printing costs $51 more per collector than molding.

Here, too, is the answer to a common question: how much does it cost to make a prototype? The prototype itself is the small number. The tooling that follows it is the deciding factor.

Step 6: pilot run and production

Before the full order, the factory makes a small batch called a pilot run, typically 30 to 300 units depending on the process. The design was settled in step 4. Now the factory is the one being tested. Do the parts fit together on the line, and does the hundredth unit match the first? Even the box gets checked against the product.

A problem found in a pilot run costs a few units. The same problem found after shipping is in 5,000 boxes.

Then production begins. From here the work is checking. Parts are inspected as they arrive. Finished units are tested at random. Every change goes on record.

Limits of this example

• All costs and the $250 price target are assumptions chosen to show the arithmetic. A real project gets real quotes.
• The squash ball collector is not developed past the concept stage.
• Electronics, software, and safety certification are not covered. A push roller needs none of them. Many sports products need all three.

Common questions

Do I need a patent before I talk to a designer? No. A non-disclosure agreement protects the conversation. Whether to file a patent is a separate question for a patent attorney.

Can I skip the proof of concept? Rarely. It is the cheapest step. Every later step depends on the answer it gives.

Should one machine serve two sports? Not at first. Build for the sport with the clearest problem, then adapt.

How long does the whole process take? It can take anywhere between 2 months and 2 years. It depends on your requirements, budget, and grit.

Where we fit

Sources: NOCSAE ND049-19, standard performance specification for newly manufactured lacrosse balls · Squash ball specification, as summarised by Wikipedia · Random close packing of spheres · Kollectaball K-Field ball collector · Protolabs: injection molding pricing · Xometry: injection molding cost · Fictiv: pilot production · StudioRed: product design cost · Cad Crowd: design and prototyping rates. Related: how much a hardware prototype costs · how to turn an idea into a product.