Kit parts should build most of your FIRST Tech Challenge robot. Custom machining earns its place on the final tenth, where no off-the-shelf part hits the exact shape a game demands. The real question during build season is not which is better in the abstract. It is which one protects your iteration speed while you still have weeks left to improve.

Kit parts win because they let you rebuild fast

Kit ecosystems such as goBILDA, REV and TETRIX share standard hole patterns, channel and brackets. That standardisation is the whole point. You can unbolt a subsystem, move a motor two holes over, and be testing again before the next practice match. Every change is reversible, and nothing is lost when an idea does not work out.

During build season your constraint is rarely machining skill. It is how many design loops you can run before your first qualifier. A team that reruns its intake ten times in kit form will usually beat a team that machined one beautiful version and then cannot easily change it. Kit parts keep that loop cheap and fast.

Where custom machining actually earns its place

Custom parts pay off on the last ten percent of the robot, and almost never before you have proven the idea in kit form. The good candidates are game-specific shapes that stock parts cannot make: an intake mouth at a particular angle, a curved guide that funnels a game element, a low-profile bracket that clears a bumper, or weight taken out of a part you have already validated.

The rule that keeps teams out of trouble is simple. Prototype in kit, commit in metal. If you have not driven the mechanism and watched it score, you are not ready to machine it. Cutting a design you have not tested just makes an expensive version of a guess.

The cost and lead-time math

The honest comparison is not really about part quality. A machined bracket can be lighter and stronger. It is about how each choice affects the two things you never have enough of: time and revisions.

Factor	Kit parts	Custom machining
Time to first working version	Minutes to an hour	Days, once design and queue are added
Cost per change	Low; reuse and re-drill existing stock	Higher; each revision is new material and shop time
Lead time	Already in your parts bin	Design, machine or outsource, then finish
Reversibility	Unbolt and try again	Committed once it is cut
Best used for	Structure, drivetrain, most brackets	Proven, game-specific geometry

Read the table as a workflow, not a verdict. Most of the robot lives in the left column all season. A few proven parts graduate to the right column once the design has stopped changing.

What to customise first

If you only machine one thing, make it the intake. Scoring usually depends on how reliably you can collect and control game elements, and those elements change shape and size every season, so stock geometry rarely fits perfectly. Intake geometry, meaning the mouth angle, the gap and the compliance of the surface that touches the element, is where a custom part turns a robot that mostly works into one that works every cycle.

After the intake, the next candidates are the parts that must survive contact or fit a tight packaging constraint. Leave drivetrain, structure and most mounting to the kit. Those are solved problems, and reinventing them costs the season time you would rather spend driving.

How to get parts made without a mill

You do not need a machine shop in your classroom to run custom parts. A few access routes cover most teams:

• Outsourced CNC. Send a shop a DXF or STEP file and expect a queue. Order early, batch your parts, and design so a single setup makes several at once.
• 3D printing as a bridge. Print non-structural geometry to iterate the shape cheaply, then machine the final version only if it has to carry load or take impact.
• Hand tools and a drill press. Many custom parts are just slots, holes and a cut edge. A file, a hacksaw and careful marking out go further than people expect.
• Makerspaces, sponsors and universities. A local fab lab or an engineering department can often cut a part faster than you can buy a mill, and the relationship is worth more than the part.

A South African reality check

Two local factors change the math. Kit parts are often imported, so lead times can run into weeks; order spares before the season, not during it. Load shedding can stall a long print or a shop schedule without warning, so treat any single machined part as a risk until it is in your hand. Keep a spares bin of the brackets and channel you use most, and buy common wear items in pairs.

Skills matter more than any single part. A team that can model a bracket, read a drawing and reason about a mechanism will make better custom-versus-kit calls all season. If you are still sourcing parts, our store is one starting point for kits and spares, and our robotics academy in Cape Town builds the design fundamentals behind these decisions. The school-holiday workshops are a low-pressure place to practise prototyping, and younger students often start with the sheenbot∞ board to learn control logic before they touch a competition robot.

Takeaway

Build the robot from kit parts, protect your iteration speed, and machine only what you have already proven, starting with the intake. Custom work is a finishing move, not a starting position. The teams that win this trade-off are the ones that keep changing their minds cheaply for as long as possible, then commit metal to the few decisions that have stopped changing.

Common questions

Should a rookie team machine anything in year one?

Usually not. Spend your first season learning to iterate in kit form and to drive well. You will learn more from ten quick rebuilds than from one machined part, and you will make far better custom choices next year.

Should we 3D print or CNC a custom intake?

Print to find the shape, machine to keep it. A printed intake is ideal for testing the mouth angle and gap over a weekend. Once you know the geometry works and it has to survive repeated contact, that is the moment to cut it in metal.

How do we build design skills before the season starts?

Practise off-season on low-stakes builds. Rehearsing modelling, prototyping and the prototype-then-commit habit before the real clock is running is what separates teams that customise well from teams that customise too soon.