Engineering Design Process for Students | Advanced Science Tutorials

The engineering design process for students is a structured way to solve practical problems using evidence. It usually begins by defining a need, understanding criteria and constraints, generating more than one possible solution, planning, building or modelling, testing, analysing results and improving the design. Unlike a one-shot craft activity, engineering expects iteration: a design that fails a test has produced useful information about what should change next.

This Advanced Science Tutorials guide is written for parents and students in Sengkang, Punggol and across Singapore who search for engineering design process, STEM engineering, ask imagine plan create improve, design challenge, prototype, engineering project, criteria and constraints and STEM activities for students. It connects Primary curiosity, PSLE scientific reasoning and Secondary G1, G2 and G3 problem solving without turning every activity into a competition to build the flashiest object.

NC State’s Engineering Place describes engineering design as an iterative process that can move through Ask, Imagine, Plan, Create, Evaluate and Share, while other education frameworks use closely related labels such as Ask, Imagine, Plan, Create, Test and Improve. The exact labels vary; the underlying logic is stable: define the problem, generate options, test against criteria, and improve using evidence. See NC State: Engineering Design Process and University of Arkansas STEM Education: Engineering Design Process.

Engineering in one sentence

Engineering uses Science, Mathematics, evidence and design judgment to create or improve products, processes and systems that meet defined needs under real constraints.

The core engineering cycle

Ask: define the need or problem.
Research: understand users, context, Science and existing solutions.
Imagine: generate multiple possibilities.
Plan: choose a concept and make the design explicit.
Create: build a prototype, model or simulation.
Test: measure performance against criteria.
Evaluate: analyse evidence and failure modes.
Improve: change the design deliberately.
Communicate: document decisions, evidence and trade-offs.
Repeat until the design is good enough for the intended purpose.

Safety first

Student engineering should use age-appropriate materials and controlled energy. Avoid mains electricity, high-speed projectiles, pressurised devices, flames, hazardous chemicals, cutting tools without supervision, structural loads that can injure, and any design that could harm people or animals. Use low-voltage educational kits, lightweight prototypes, simulations and teacher-approved materials.

An engineering challenge does not become more authentic because it is dangerous. Real engineering is defined partly by controlling risk.

Problem definition

Core idea. A useful design problem states what must be improved or achieved and who needs it.

Common mistake. Students often begin building before agreeing on the actual problem.

Design evidence. A classroom device that keeps pencils from rolling off a desk has a clearer need than ‘build something cool’.

A student should connect this idea to criteria and constraints. Ask what requirement it protects, what could be measured, and what trade-off might appear. Engineering becomes more rigorous when the learner can explain why a design choice exists instead of only describing what was built.

For Primary students, use lightweight structures, paper, cardboard, safe craft materials and simple measurements. For Secondary G1, G2 and G3 learners, add graphs, calculations, simulation, reliability and more explicit trade-offs. Keep every prototype within safe energy, load and material limits.

Transfer check: give the learner a different design problem and ask where the same principle appears. If the idea travels from bridges to insulation, circuits, packaging or digital systems, it is becoming engineering knowledge rather than one-project memory.

User need

Core idea. Engineering begins with a person, organisation or system that has a need.

Common mistake. Students may optimise for what they enjoy building rather than what the user requires.

Design evidence. Interview questions, observation or a written brief can reveal what matters to the intended user.

Criteria

Core idea. Criteria describe what successful performance looks like.

Common mistake. Students often say ‘works well’ without defining measurable success.

Design evidence. A bridge might need to span 30 cm and support 500 g without excessive bending.

Constraints

Core idea. Constraints limit acceptable solutions: cost, size, materials, time, safety, energy or environmental impact.

Common mistake. Students may treat constraints as annoying rules rather than part of the design problem.

Design evidence. A budget or size limit forces trade-offs and creativity.

Research

Core idea. Research identifies prior solutions, relevant Science, materials and known failure modes.

Common mistake. Students sometimes copy an existing design without understanding why it works.

Design evidence. Looking at several bridge types can reveal load paths without requiring imitation of one answer.

Multiple concepts

Core idea. Good design usually begins with more than one idea.

Common mistake. Students lock onto the first idea and defend it before testing alternatives.

Design evidence. Sketching three concepts creates options to compare against criteria.

Brainstorming

Core idea. Brainstorming separates idea generation from early judgment.

Common mistake. Students self-censor unusual ideas too quickly or copy the loudest teammate.

Design evidence. Time-limited independent idea generation before group discussion protects diversity of thinking.

Sketching

Core idea. Sketches externalise geometry, components and relationships before building.

Common mistake. Students draw decorative pictures without dimensions or labels.

Design evidence. A functional sketch should show dimensions, interfaces and important materials.

Decision matrix

Core idea. A decision matrix compares options against weighted criteria.

Common mistake. Students choose the design they personally like rather than the one best aligned to the brief.

Design evidence. Scoring strength, cost, mass and ease of assembly makes trade-offs explicit.

Prototype

Core idea. A prototype is a representation built to learn about a design, not necessarily a polished final product.

Common mistake. Students treat the first build as the final answer.

Design evidence. A rough paper model can reveal geometry problems before expensive materials are used.

Model

Core idea. A model simplifies selected features of a system.

Common mistake. Students expect every model to behave exactly like the final product.

Design evidence. A cardboard bridge can test shape and load path while failing to represent weather resistance.

Simulation

Core idea. A simulation uses rules or equations to explore system behaviour.

Common mistake. Students may treat simulated output as guaranteed real-world performance.

Design evidence. Simulation results depend on assumptions and should be checked against physical evidence where possible.

Test plan

Core idea. A test plan defines what will be measured, how, under which conditions and against what pass criterion.

Common mistake. Students build first and invent the test after seeing what the prototype can do.

Design evidence. Writing the test before building prevents moving the goalposts.

Baseline

Core idea. A baseline shows current or reference performance.

Common mistake. Students claim improvement without measuring what existed before.

Design evidence. Testing the original design before modifying it allows meaningful comparison.

Measurement

Core idea. Engineering tests need quantities and units tied to criteria.

Common mistake. Students use vague judgments such as ‘stronger’ or ‘faster’.

Design evidence. Measure supported mass, deflection, time, temperature, distance or another relevant variable.

Repeatability

Core idea. Repeated tests reveal whether performance is consistent.

Common mistake. Students celebrate one successful run and stop.

Design evidence. Multiple trials can show whether a design succeeds reliably or only occasionally.

Failure mode

Core idea. A failure mode describes how a design stops meeting its requirement.

Common mistake. Students call the whole design a failure without identifying where or how it failed.

Design evidence. A bridge may buckle at a joint, twist laterally or exceed allowed deflection before collapse.

Root cause

Core idea. Root-cause analysis asks why the failure occurred rather than treating the visible symptom as the cause.

Common mistake. Students reinforce the broken part without understanding why load concentrated there.

Design evidence. Tracing force paths or component interactions can reveal a deeper cause.

Iteration

Core idea. Iteration means deliberately changing a design and testing again.

Common mistake. Students rebuild randomly after failure.

Design evidence. One controlled change at a time makes evidence about improvement easier to interpret.

Trade-off

Core idea. A trade-off occurs when improving one criterion makes another worse.

Common mistake. Students expect one design to maximise every property simultaneously.

Design evidence. A thicker structure may be stronger but heavier and more expensive.

Optimisation

Core idea. Optimisation seeks the best performance under defined criteria and constraints.

Common mistake. Students confuse ‘best’ with one universal design.

Design evidence. The optimal solution depends on how criteria are weighted.

Robustness

Core idea. A robust design works despite ordinary variation in conditions.

Common mistake. Students test only one perfect setup.

Design evidence. Varying load position or user handling can reveal fragile designs.

Reliability

Core idea. Reliability concerns consistent performance over repeated use or time.

Common mistake. Students assume a prototype that works once is ready.

Design evidence. Repeated cycles can reveal wear, loosening or drift.

Tolerance

Core idea. Tolerance defines acceptable variation in dimensions or performance.

Common mistake. Students assume every component can be made exactly to one value.

Design evidence. Manufacturing variation means designs need room for small differences.

Factor of safety

Core idea. A safety factor provides margin between expected load and failure capacity.

Common mistake. Students design exactly at expected maximum load.

Design evidence. Extra margin helps account for uncertainty and unexpected conditions.

Materials selection

Core idea. Material choice should follow required properties, availability, cost and safety.

Common mistake. Students choose materials by appearance or familiarity.

Design evidence. A beam may need stiffness; insulation needs low thermal conductivity; wiring needs conductivity and safe insulation.

Structures

Core idea. Structures carry loads through members, joints and supports.

Common mistake. Students focus on total material amount rather than geometry.

Design evidence. Triangles, folds and load paths can increase stiffness without simply adding mass.

Tension and compression

Core idea. Structural members may carry pulling or pushing forces.

Common mistake. Students think every beam experiences only one type of stress.

Design evidence. A truss distributes tension and compression through different members.

Bending

Core idea. Bending combines tension and compression across a member.

Common mistake. Students add material anywhere instead of where geometry increases stiffness.

Design evidence. Changing cross-section depth can strongly affect bending performance.

Buckling

Core idea. Long slender members can fail by instability under compression.

Common mistake. Students assume failure occurs only when material ‘breaks’.

Design evidence. A straw column can bend sideways well before the material crushes.

Joints

Core idea. Connections often control system strength and ease of assembly.

Common mistake. Students focus on strong components while ignoring weak joints.

Design evidence. A bridge with excellent beams can fail where tape or glue transfers load poorly.

Mechanisms

Core idea. Mechanisms transform motion and force using links, gears, cams, pulleys or levers.

Common mistake. Students treat moving parts as decoration.

Design evidence. A mechanism should be analysed by input, transformation and output.

Levers

Core idea. Levers trade force and movement through pivot geometry.

Common mistake. Students memorise lever classes without seeing mechanical advantage.

Design evidence. Changing distances from the pivot changes required input force.

Gears

Core idea. Gears transfer rotation and can trade speed for torque.

Common mistake. Students think larger gears are always ‘more powerful’.

Design evidence. Gear ratio predicts relative rotational speeds and torque trends.

Pulleys

Core idea. Pulley systems can redirect force and provide mechanical advantage.

Common mistake. Students assume adding any pulley automatically reduces input force.

Design evidence. Only certain arrangements create mechanical advantage.

Energy efficiency

Core idea. Efficiency compares useful output with energy input.

Common mistake. Students assume no energy losses if a device works.

Design evidence. Friction, electrical resistance and thermal losses reduce useful output.

Thermal design

Core idea. Thermal engineering controls heat transfer through conduction, convection and radiation.

Common mistake. Students choose insulation without defining the thermal goal.

Design evidence. A container designed to retain heat and one designed to dissipate heat require opposite strategies.

Electrical design

Core idea. Electrical engineering requires safe voltage, correct connections, component ratings and protection.

Common mistake. Students treat wires as interchangeable lines with no resistance or safety limits.

Design evidence. Low-voltage educational circuits can demonstrate series, parallel and control logic safely.

Sensors

Core idea. Sensors convert physical conditions into measurable signals.

Common mistake. Students assume a sensor reading is direct truth.

Design evidence. Calibration, range, placement and response time affect measurements.

Feedback

Core idea. Feedback uses measured output to influence system behaviour.

Common mistake. Students think control means a human constantly adjusting the system.

Design evidence. A thermostat compares temperature with a set point and switches heating or cooling accordingly.

Coding

Core idea. Software can control devices, process sensor data and automate decisions.

Common mistake. Students treat code as separate from engineering hardware.

Design evidence. A robot’s physical design and control algorithm interact as one system.

Systems thinking

Core idea. A system contains components whose interactions create overall behaviour.

Common mistake. Students optimise one component while harming the whole system.

Design evidence. A lighter battery may reduce mass but shorten operating time.

Interfaces

Core idea. Interfaces are boundaries where components, users or systems interact.

Common mistake. Students design components independently and discover late that they do not fit.

Design evidence. Connector shape, data format and user controls are interface decisions.

Ergonomics

Core idea. Ergonomics considers human dimensions, comfort, reach and workload.

Common mistake. Students design for themselves only.

Design evidence. Testing with representative users can reveal controls that are too small or awkward.

Accessibility

Core idea. Accessible design reduces unnecessary barriers for users with different abilities.

Common mistake. Students treat accessibility as an optional add-on.

Design evidence. Contrast, tactile cues, captions and alternative controls can expand usability.

Sustainability

Core idea. Sustainable design considers materials, energy, waste, durability, repair and end-of-life impacts.

Common mistake. Students equate sustainability with one recyclable material.

Design evidence. Whole-system trade-offs matter across the life cycle.

Life-cycle thinking

Core idea. Life-cycle analysis considers extraction, manufacture, transport, use and disposal.

Common mistake. Students judge environmental impact only from the use phase.

Design evidence. A low-energy product may still have high manufacturing or disposal impacts.

Cost

Core idea. Cost includes materials, manufacturing, maintenance, energy and failure risk.

Common mistake. Students compare only purchase price.

Design evidence. A more durable design can cost more initially but less over its lifetime.

Maintainability

Core idea. Maintainability concerns how easily a system can be inspected, serviced or repaired.

Common mistake. Students seal everything permanently because it looks neat.

Design evidence. Accessible fasteners and modular parts can reduce repair cost.

Modularity

Core idea. Modular designs divide systems into replaceable or upgradeable units.

Common mistake. Students assume one-piece construction is always simpler.

Design evidence. Modules can improve repair and reuse but add interfaces and connectors.

Documentation

Core idea. Engineering documentation records requirements, drawings, tests and decisions.

Common mistake. Students see documentation as paperwork after the build.

Design evidence. Good records let teams reproduce, diagnose and improve designs.

Version control

Core idea. Version control distinguishes one design iteration from another.

Common mistake. Students change several things and forget what produced improvement.

Design evidence. Labelled versions connect modifications to test results.

Teamwork

Core idea. Engineering teams combine different knowledge and perspectives.

Common mistake. Students divide work into isolated tasks and never integrate.

Design evidence. Shared criteria, interfaces and test evidence keep collaboration coherent.

Technical communication

Core idea. Engineers communicate through drawings, specifications, graphs, reports and presentations.

Common mistake. Students decorate presentations without showing evidence.

Design evidence. A clear graph of performance versus criterion often communicates more than a slogan.

Ethics

Core idea. Engineering decisions can affect safety, privacy, environment and fairness.

Common mistake. Students think meeting a technical target is enough.

Design evidence. A device that performs well but creates unacceptable risk is not a successful design.

Risk assessment

Core idea. Risk combines likelihood and consequence of harm.

Common mistake. Students focus only on dramatic hazards.

Design evidence. Low-probability severe failures and frequent minor failures both deserve analysis.

Redundancy

Core idea. Redundancy adds backup pathways or components to improve resilience.

Common mistake. Students think duplicated components are always wasteful.

Design evidence. Critical systems may justify redundancy because failure consequences are high.

Graceful failure

Core idea. A good design may fail in a controlled, detectable way rather than catastrophically.

Common mistake. Students judge only whether failure occurs.

Design evidence. A fuse that safely interrupts current can be a successful failure-management feature.

Testing to limits

Core idea. Limit testing can reveal margins and failure modes when done safely.

Common mistake. Students push prototypes dangerously to destruction without purpose.

Design evidence. Use low-energy models or simulations and define stop conditions before the test.

Engineering and Science

Core idea. Science asks what happens and why; engineering asks how to create a solution under constraints.

Common mistake. Students treat them as unrelated school subjects.

Design evidence. Engineering uses scientific models, while engineering problems can generate new scientific questions.

Engineering and Mathematics

Core idea. Mathematics quantifies dimensions, loads, rates, costs and uncertainty.

Common mistake. Students use formulas without relating them to design choices.

Design evidence. A calculation is valuable when it changes a decision.

Engineering and computing

Core idea. Computing supports modelling, control, optimisation, data analysis and design automation.

Common mistake. Students think programming is useful only for software products.

Design evidence. Code increasingly shapes physical systems from sensors to transportation.

Engineering careers

Core idea. Engineering includes civil, mechanical, electrical, chemical, environmental, biomedical, aerospace, software and many hybrid fields.

Common mistake. Students think ‘engineer’ means one kind of builder.

Design evidence. Different disciplines share problem definition, evidence and iteration while using different technical knowledge.

Ten safe engineering challenges

Build a paper bridge that spans a fixed gap and supports a defined light load.
Design a foil boat that holds the greatest safe mass of identical coins before water enters.
Create a paper tower that reaches a height criterion with a material limit.
Design an insulated container for safely warm water and minimise temperature change.
Create a paper helicopter that falls slowly from standing height.
Design packaging that protects a lightweight fragile model from a short tabletop drop.
Build a straw structure that resists a gentle fan-produced wind.
Design a cardboard phone stand that meets angle, stability and material constraints.
Use a simulation to optimise a low-voltage circuit without using mains electricity.
Create a water-saving process redesign using measured or published household-use data rather than modifying plumbing.

A twelve-week Engineering Design programme

Week 1: problem definition, users, criteria and constraints.
Week 2: research and multiple concepts.
Week 3: sketching, modelling and decision matrices.
Week 4: prototypes and test plans.
Week 5: measurement, baselines and repeatability.
Week 6: structures, loads and joints.
Week 7: mechanisms and energy.
Week 8: thermal and electrical systems.
Week 9: sensors, feedback and coding.
Week 10: reliability, robustness and risk.
Week 11: sustainability, cost and maintainability.
Week 12: complete design cycle, communication and reflection.

How parents can support engineering

Parents can help define safe materials, time and budget constraints, but should not quietly rebuild a child’s prototype after it fails. Ask what the test showed, which criterion was missed, and what one change will be tested next. A failed prototype with a good diagnosis is often more educational than a perfect adult-built model.

How tutors can use a three-student design lesson

Give all three students the same problem but require independent concepts before discussion. Compare which criteria each student prioritised, how they interpreted constraints and what test would discriminate among designs. Then let the test evidence, not confidence, drive the next iteration.

Frequently asked questions

What are the steps of the engineering design process?

Common versions include Ask, Imagine, Plan, Create, Test and Improve, with research, evaluation and communication often included. The process is iterative rather than rigidly linear.

How is engineering different from Science?

Science focuses on explaining and predicting natural phenomena; engineering focuses on creating solutions under criteria and constraints. They rely heavily on each other.

What makes a good student engineering project?

A clear problem, measurable criteria, real constraints, multiple concepts, a safe prototype, a defined test and evidence-based iteration.

Does a prototype have to work the first time?

No. A prototype is partly a learning tool. A failed test is useful when it reveals why the design missed a criterion.

Should students optimise one thing at a time?

Often yes during diagnosis, but real designs involve trade-offs across several criteria. Controlled changes make evidence easier to interpret.

Can engineering be taught before Secondary school?

Yes. Young students can define needs, build safe prototypes, test and improve without advanced Mathematics or hazardous tools.

Final operating rule

Engineering starts with a need and ends only when evidence says the solution is good enough for its purpose. Define the problem. Make criteria measurable. Respect constraints. Generate options. Prototype safely. Test honestly. Diagnose failure. Improve deliberately. Communicate the trade-offs. Then repeat. The engineering design process is not a craft recipe; it is a disciplined way to turn Science and Mathematics into useful decisions.

Problem definition — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where problem definition appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students often begin building before agreeing on the actual problem.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A useful design problem states what must be improved or achieved and who needs it.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A classroom device that keeps pencils from rolling off a desk has a clearer need than ‘build something cool’.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Finish with iteration. Change one high-leverage feature, test again, and compare the new data with the prior version. Parents should protect safety and constraints; tutors should require independent concepts before discussion. The final proof is a student who can justify why version two is better rather than merely saying it looks better.

User need — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where user need appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students may optimise for what they enjoy building rather than what the user requires.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Engineering begins with a person, organisation or system that has a need.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Interview questions, observation or a written brief can reveal what matters to the intended user.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Criteria — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where criteria appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students often say ‘works well’ without defining measurable success.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Criteria describe what successful performance looks like.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A bridge might need to span 30 cm and support 500 g without excessive bending.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Constraints — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where constraints appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students may treat constraints as annoying rules rather than part of the design problem.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Constraints limit acceptable solutions: cost, size, materials, time, safety, energy or environmental impact.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A budget or size limit forces trade-offs and creativity.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Research — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where research appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students sometimes copy an existing design without understanding why it works.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Research identifies prior solutions, relevant Science, materials and known failure modes.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Looking at several bridge types can reveal load paths without requiring imitation of one answer.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Multiple concepts — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where multiple concepts appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students lock onto the first idea and defend it before testing alternatives.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Good design usually begins with more than one idea.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Sketching three concepts creates options to compare against criteria.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Brainstorming — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where brainstorming appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students self-censor unusual ideas too quickly or copy the loudest teammate.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Brainstorming separates idea generation from early judgment.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Time-limited independent idea generation before group discussion protects diversity of thinking.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Sketching — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where sketching appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students draw decorative pictures without dimensions or labels.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Sketches externalise geometry, components and relationships before building.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A functional sketch should show dimensions, interfaces and important materials.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Decision matrix — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where decision matrix appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students choose the design they personally like rather than the one best aligned to the brief.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A decision matrix compares options against weighted criteria.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Scoring strength, cost, mass and ease of assembly makes trade-offs explicit.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Prototype — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where prototype appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students treat the first build as the final answer.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A prototype is a representation built to learn about a design, not necessarily a polished final product.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A rough paper model can reveal geometry problems before expensive materials are used.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Model — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where model appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students expect every model to behave exactly like the final product.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A model simplifies selected features of a system.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A cardboard bridge can test shape and load path while failing to represent weather resistance.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Simulation — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where simulation appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students may treat simulated output as guaranteed real-world performance.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A simulation uses rules or equations to explore system behaviour.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Simulation results depend on assumptions and should be checked against physical evidence where possible.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Test plan — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where test plan appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students build first and invent the test after seeing what the prototype can do.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A test plan defines what will be measured, how, under which conditions and against what pass criterion.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Writing the test before building prevents moving the goalposts.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Baseline — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where baseline appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students claim improvement without measuring what existed before.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A baseline shows current or reference performance.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Testing the original design before modifying it allows meaningful comparison.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Measurement — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where measurement appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students use vague judgments such as ‘stronger’ or ‘faster’.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Engineering tests need quantities and units tied to criteria.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Measure supported mass, deflection, time, temperature, distance or another relevant variable.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Repeatability — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where repeatability appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students celebrate one successful run and stop.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Repeated tests reveal whether performance is consistent.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Multiple trials can show whether a design succeeds reliably or only occasionally.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Failure mode — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where failure mode appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students call the whole design a failure without identifying where or how it failed.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A failure mode describes how a design stops meeting its requirement.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A bridge may buckle at a joint, twist laterally or exceed allowed deflection before collapse.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Root cause — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where root cause appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students reinforce the broken part without understanding why load concentrated there.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Root-cause analysis asks why the failure occurred rather than treating the visible symptom as the cause.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Tracing force paths or component interactions can reveal a deeper cause.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Iteration — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where iteration appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students rebuild randomly after failure.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Iteration means deliberately changing a design and testing again.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: One controlled change at a time makes evidence about improvement easier to interpret.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Trade-off — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where trade-off appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students expect one design to maximise every property simultaneously.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A trade-off occurs when improving one criterion makes another worse.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A thicker structure may be stronger but heavier and more expensive.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Optimisation — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where optimisation appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students confuse ‘best’ with one universal design.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Optimisation seeks the best performance under defined criteria and constraints.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: The optimal solution depends on how criteria are weighted.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Robustness — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where robustness appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students test only one perfect setup.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A robust design works despite ordinary variation in conditions.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Varying load position or user handling can reveal fragile designs.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Reliability — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where reliability appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students assume a prototype that works once is ready.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Reliability concerns consistent performance over repeated use or time.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Repeated cycles can reveal wear, loosening or drift.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Tolerance — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where tolerance appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students assume every component can be made exactly to one value.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Tolerance defines acceptable variation in dimensions or performance.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Manufacturing variation means designs need room for small differences.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Factor of safety — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where factor of safety appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students design exactly at expected maximum load.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—A safety factor provides margin between expected load and failure capacity.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Extra margin helps account for uncertainty and unexpected conditions.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Materials selection — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where materials selection appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students choose materials by appearance or familiarity.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Material choice should follow required properties, availability, cost and safety.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A beam may need stiffness; insulation needs low thermal conductivity; wiring needs conductivity and safe insulation.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Structures — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where structures appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students focus on total material amount rather than geometry.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Structures carry loads through members, joints and supports.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Triangles, folds and load paths can increase stiffness without simply adding mass.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Tension and compression — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where tension and compression appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students think every beam experiences only one type of stress.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Structural members may carry pulling or pushing forces.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A truss distributes tension and compression through different members.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Bending — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where bending appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students add material anywhere instead of where geometry increases stiffness.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Bending combines tension and compression across a member.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Changing cross-section depth can strongly affect bending performance.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Buckling — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where buckling appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students assume failure occurs only when material ‘breaks’.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Long slender members can fail by instability under compression.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A straw column can bend sideways well before the material crushes.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Joints — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where joints appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students focus on strong components while ignoring weak joints.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Connections often control system strength and ease of assembly.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A bridge with excellent beams can fail where tape or glue transfers load poorly.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Mechanisms — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where mechanisms appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students treat moving parts as decoration.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Mechanisms transform motion and force using links, gears, cams, pulleys or levers.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A mechanism should be analysed by input, transformation and output.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Levers — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where levers appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students memorise lever classes without seeing mechanical advantage.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Levers trade force and movement through pivot geometry.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Changing distances from the pivot changes required input force.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Gears — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where gears appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students think larger gears are always ‘more powerful’.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Gears transfer rotation and can trade speed for torque.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Gear ratio predicts relative rotational speeds and torque trends.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Pulleys — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where pulleys appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students assume adding any pulley automatically reduces input force.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Pulley systems can redirect force and provide mechanical advantage.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Only certain arrangements create mechanical advantage.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Energy efficiency — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where energy efficiency appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students assume no energy losses if a device works.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Efficiency compares useful output with energy input.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Friction, electrical resistance and thermal losses reduce useful output.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Thermal design — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where thermal design appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students choose insulation without defining the thermal goal.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Thermal engineering controls heat transfer through conduction, convection and radiation.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: A container designed to retain heat and one designed to dissipate heat require opposite strategies.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Electrical design — design clinic 1

Start with a new engineering brief unrelated to the original example. Ask where electrical design appears in the problem and which criterion or constraint makes it relevant. The learner must connect the concept to a measurable design decision rather than recite a definition.

Surface the failure mode: Students treat wires as interchangeable lines with no resistance or safety limits.. Ask what would happen if a team followed that mistaken approach. Then use the better principle—Electrical engineering requires safe voltage, correct connections, component ratings and protection.—to redesign the decision. Record the prediction before building or simulating so the test can genuinely confirm or challenge the idea.

Use the evidence route: Low-voltage educational circuits can demonstrate series, parallel and control logic safely.. Decide what variable will be measured, what baseline will be used and what result counts as improvement. If the test does not distinguish the old and new designs, the test plan is too weak and should be revised before resources are spent.

Small Group Tutorials

Advanced Science Tutorials | Engineering Design Process for Students: Ask, Imagine, Plan, Create, Test and Improve