Design process

The engineering design process uses requirements, alternative designs, prototypes, and tests to decide what to build. Each revision should answer a question about the design.

Design requirements

"Build an intake" already assumes a solution. "Acquire the object while the robot approaches with imperfect alignment" identifies a problem several mechanisms might solve.

Separate constraints, which the design must meet, from targets, where you can accept a compromise. Size limits, workshop tools, and a fixed deadline are constraints. Pickup time, mass, and miss rate are targets until the team decides on a firm limit.

The specific target values should come from your strategy and measurements.

Concept comparison

Sketch concepts that differ in how they constrain or move the object. Three cosmetic versions of one claw do not investigate whether a roller intake would be better.

Set aside designs that cannot meet a constraint. Compare the others on pickup tolerance, size, build time, and access for repairs. If a score is the team’s judgment rather than a test result, say so.

Keep the reason for rejecting an option. That record prevents the same unworkable idea returning when a new student joins.

Prototyping

A prototype is a tool for answering a question. Use a cardboard envelope to check packaging, an adjustable roller mount to find compression, or a simple shaft fixture to test load. It need not resemble the final robot if it preserves the physics under investigation.

For example, an intake experiment might vary roller gap while keeping roller material, speed, object, approach path, and fixture constant. Test several object orientations at each gap. Record capture, rejection, damage, and recovery; do not record only successful attempts.

CAD is valuable for geometry and interference. Physical tests reveal friction, compliance, wear, and manufacturing variation. Neither alone establishes full robot performance.

Design decisions

Before a test, write down what you expect, what you will change, what you will measure, and when to stop. Afterward, compare the result with your target. Keep what you saw separate from why you think it happened: “the object rotated out of the gap” is an observation; “the contact force was too high” is a possible explanation.

If the explanation is uncertain, design the next experiment to distinguish causes. Reducing speed, changing compression, and replacing the roller simultaneously may improve pickup without revealing which change helped.

Use this compact record:

Question: Which roller gap captures the chosen range of object orientations?
Revision: intake fixture A, roller material and dimensions recorded
Variable: gap at the contact region
Controlled conditions: speed command, approach fixture, object set
Measurements: attempts, captures, acquisition time, damage, recovery
Decision: selected geometry and reason, or further test needed
Evidence: dated photos, video, and raw trial table

Integration testing

A mechanism that passes on a bench can fail when chassis flex changes its alignment or another mechanism consumes the same space. Repeat the relevant test on the assembled robot, with the intended payload and simultaneous motion.

For each connection between mechanisms, record the mounting references, allowed flex, space for movement, power demand, sensor readings, and which code controls it. If one changes, repeat the tests that rely on it.