01 / 20

Improve Device
Design

Learn how to use test results to modify and optimize a device's performance through the engineering design process.

Physical Science

What Is the Engineering Design Process?

The engineering design process is a series of steps engineers follow to come up with a solution to a problem. Unlike a recipe, the process is iterative: you cycle back through steps to keep improving your design.

Define Research Prototype Test Improve Redesign

Key Terms You Need to Know

Prototype
A working model of a device built to test ideas before making the final version.
Variable
A factor you change, measure, or keep the same during a test.
Constraint
A limit on your design, such as time, cost, size, or available materials.
Criteria
The requirements your device must meet to be considered successful.
Iteration
One cycle of designing, testing, and improving. More iterations = better design.
Optimization
The process of making a design as effective as possible within constraints.

Why Do Test Results Matter?

Without testing, you are just guessing whether your device works well. Test results give you evidence: real numbers and observations that tell you what is working and what needs to change.

Before Testing "I think it works" "It looks right" "Maybe it's fast enough" = Guessing After Testing "It held 2.5 kg" "Speed: 1.8 m/s" "Failed at joint B" = Evidence

The Core Rule of Optimization

✕ Multiple Changes Changed material Changed shape Changed size Result improved... but WHY? Unknown! ✓ One Change at a Time Changed material only Shape: same Size: same Result improved! Cause: new material ✓

The Optimization Cycle

1. Test 2. Collect Data 3. Analyze 4. Identify Weakness 5. Modify 6. Re-test

Testing vs. Improving

Testing and improving are different phases of the same cycle. Many students confuse them, but each has a unique purpose.

Testing Improving
Goal Gather data on performance Make changes based on data
Question "How well does it work?" "How can I make it better?"
Output Numbers, observations, notes A modified prototype
When Before any change is made After data is analyzed

How to Apply: Step-by-Step

1
State your criteria. What does success look like? Write measurable goals (e.g., "hold at least 3 kg").
2
Test your prototype. Run multiple trials and record all data carefully.
3
Analyze results. Compare data to your criteria. Did it meet the goal? Where did it fall short?
4
Identify the weakest point. What specific part or feature caused the biggest gap between results and criteria?
5
Modify one variable. Change only the weakest element. Keep everything else the same.
6
Re-test and compare. Did the change improve performance? Repeat the cycle.

Example 1: The Paper Bridge

A student builds a paper bridge to hold weight. The criterion: the bridge must hold at least 500 g without collapsing. Here are the test results from three iterations:

0 200 400 600 800 Weight (g) Goal 250 g V1: Flat 450 g V2: Folded 700 g V3: Truss

Analyzing the Bridge Data

Let's trace the student's thinking at each iteration to see how test results guided modifications:

1
V1: Flat sheet. Collapsed at 250 g. Observation: the paper buckled in the center. Weakness: no structural support in the middle.
2
V2: Accordion fold. Held 450 g. Better, but still below goal. Observation: folds spread apart under load. Weakness: folds lack lateral stability.
3
V3: Triangular truss. Held 700 g. Exceeded goal. Triangles resist spreading because of their rigid geometry. Success.

Example 2: The Solar Oven

A team designs a solar oven. The criterion: heat a cup of water to at least 60°C in 30 minutes. They test three iterations and record peak temperature:

Version Change Made Peak Temp Met Goal?
V1 Cardboard box, no lining 38°C ✕
V2 Added aluminum foil reflector 52°C ✕
V3 Added clear plastic lid to trap heat 67°C ✓

Example 3: Egg Drop Challenge

Students must protect an egg dropped from 3 meters. Criterion: egg survives with no cracks. Constraint: total device mass < 100 g.

V1: Bubble Wrap egg Mass: 85 g Cracked ✕ No impact zone V2: + Crumple Zone egg + paper tube Mass: 92 g Survived ✓ Small crack on rim V3: + Parachute egg Mass: 98 g Perfect ✓ No damage at all

Quantifying Improvement

Engineers often measure improvement as a percent change. This lets you compare iterations no matter what units you use:

For the paper bridge example: from V1 (250 g) to V2 (450 g):

Data Tables and Fair Tests

Good test data requires a fair test. A fair test controls all variables except the one you are changing. Here is what a proper data table looks like:

Trial Variable Changed Kept Same Result
1 Blade angle: 30° Motor, weight, height 12 rotations
2 Blade angle: 45° Motor, weight, height 18 rotations
3 Blade angle: 60° Motor, weight, height 15 rotations

Common Mistakes to Avoid

✕ Changing Too Many Variables
If you change the material AND the shape at the same time, you cannot tell which change caused the result.
✕ Ignoring Data
Making changes based on feelings instead of test results. "I think it looks stronger" is not evidence.
✕ Only One Trial
A single test can be a fluke. Run at least 3 trials per version and use the average for comparison.
✕ No Baseline Measurement
Without testing V1 first, you have nothing to compare improvements against. Always start with a baseline.
✕ Forgetting Constraints
Making a device bigger or heavier to improve performance may violate size or cost constraints.
✕ Stopping Too Early
Meeting the minimum criterion once does not mean the design is optimized. Keep iterating to find the best version.

Question 1 of 5

A student's solar car travels 2 meters in 10 seconds. She wants it to go faster. She changes the wheel size AND the motor at the same time. The car now goes 2 meters in 7 seconds. What is the problem with her approach?

Question 2 of 5

A team's water filter cleans 80 mL per minute in Version 1 and 120 mL per minute in Version 2. What is the percent improvement?

Question 3 of 5

Which of the following is a constraint, not a criterion?

Question 4 of 5

A student tests a wind turbine and finds that wider blades spin slower but generate more power. She decides to make the blades even wider. After testing, power actually decreased. What most likely happened?

Question 5 of 5

A student's catapult launches a ball 4.2 m on average. After switching from a rubber band to a stronger spring, the average distance is 3.8 m. What should the student do next?