SE 54200

Welcome to SE 54200

Course Overview

• Focus: AI-accelerated rapid prototyping from concept to functional prototype
• Builds on: SE 54100 AI-Assisted Systems Engineering foundations
• Outcome: Master compressed development timelines while maintaining engineering rigor

What is Rapid Prototyping?

An iterative design methodology that compresses time from concept to physical prototype through:

• Accelerated design cycles
• Early physical realization
• Fail-fast learning approaches
• Parallel development streams

Traditional vs. Rapid Prototyping

The Prototyping Spectrum

1. Conceptual: Paper sketches, foam models
2. Functional: Working subsystems, breadboards
3. Integration: Complete system assembly
4. Pre-Production: Production-representative

AI Acceleration Opportunities

Engineering Accountability

AI Assists, You Are Responsible

"The AI suggested it" is never an acceptable answer.

Course Requirement

Creating a prototype is necessary but NOT sufficient

Students must be able to explain:

• WHY each design choice was made
• HOW each decision addresses requirements
• WHAT trade-offs were considered
• HOW AI-generated content was validated

If you cannot explain your engineering decisions, you will not pass.

Traditional Timeline

Problems: Late discovery, expensive changes, long time-to-market

AI-Accelerated Timeline

Enablers: Generative design, code generation, digital twins, AI documentation

Course Project: Smart Thermostat

with Air Quality Monitoring

• Temperature and humidity control
• Air quality monitoring (CO2, VOC)
• WiFi connectivity and voice assistant integration
• Learning algorithms for energy optimization

Thermostat Subsystems

Week 1 Exercise

Identify Acceleration Opportunities

1. What features could AI help generate or evaluate?
2. Which design tasks are candidates for generative design?
3. What components can be rapidly prototyped?
4. How can simulation reduce physical testing needs?

Deliverable: Form project teams, begin defining system concept

Week 1 Summary

• Rapid prototyping compresses timelines through iteration
• AI accelerates ideation, design, validation, and documentation
• Systems engineering provides the framework
• Smart Thermostat project demonstrates these principles

Next week: Agile/Lean prototyping methodologies

Agile for Physical Products

Scrum for Prototyping

• Sprint Duration: 2-4 weeks (longer than software)
• Sprint Goal: Demonstrable prototype increment
• Daily Standup: Cross-functional coordination
• Sprint Review: Physical prototype demonstration
• Retrospective: Process improvement

Lean Principles in Prototyping

1. Eliminate Waste: No gold-plating, MVP approach
2. Build Quality In: Test early and often
3. Create Knowledge: Document learnings
4. Defer Commitment: Keep options open

Spiral Development Model

Risk-driven iteration:

1. Determine Objectives: What are the prototype goals?
2. Identify Risks: What could go wrong?
3. Develop and Test: Build prototype to address risks
4. Plan Next Iteration: What risks remain?

Minimum Viable Prototype (MVP)

Smart Thermostat: Product Backlog

Week 2 Exercise

Create Sprint Backlog

1. Break down temperature sensing feature into tasks
2. Estimate effort for each task
3. Assign tasks to discipline areas
4. Define acceptance criteria
5. Identify dependencies between tasks

Week 2 Summary

• Agile methodologies adapt for hardware
• Lean principles reduce waste
• Spiral development addresses risks systematically
• AI accelerates iteration cycles
• MVP approach validates with minimum investment

Next week: Concept-to-Prototype Workflows

The Concept Phase

1. Problem Definition: What problem does the product solve?
2. User Research: Who are the users?
3. Competitive Analysis: What exists?
4. Concept Generation: What are possible solutions?
5. Concept Selection: Which concept to pursue?

AI-Assisted Ideation

AI tools can accelerate ideation:

• Generate feature variations
• Explore design alternatives
• Identify edge cases
• Synthesize market research

Design Thinking Process

1. Empathize: Understand user needs
2. Define: Frame the problem clearly
3. Ideate: Generate many solutions
4. Prototype: Build to learn
5. Test: Gather feedback

Concept Selection: Pugh Matrix

S = Same as baseline, + = Better, - = Worse

Thermostat Problem Statement

MoSCoW Prioritization

Week 3 Exercise

AI Ideation Session

1. Generate 10 feature ideas using AI
2. Evaluate each for feasibility and differentiation
3. Create Pugh matrix comparing top 5 concepts
4. Define 5 key requirements from selected concept

Deliverable: Concept definition document

Week 3 Summary

• Concept phase transforms ideas into requirements
• AI accelerates ideation and evaluation
• Design thinking provides structured approach
• Concept selection uses objective criteria
• Requirements must be SMART

Next week: Hardware Prototyping with AI

CAD in Prototyping

• Design visualization
• Interference checking
• Manufacturing preparation
• Documentation

Modern capabilities: Parametric modeling, cloud collaboration, generative design

Generative Design

AI-driven design exploration:

1. Define design space (envelope)
2. Specify constraints (mounting, interfaces)
3. Set optimization goals (weight, strength)
4. Select manufacturing method
5. Generate and evaluate alternatives

3D Printing Technologies

Design for 3D Printing (FDM)

• Minimum wall thickness: 1.2 mm
• Overhang angle: <45° without supports
• Bridging: <50 mm span
• Hole compensation: +0.2-0.4 mm
• Layer orientation: Consider strength direction

Rapid Prototyping Services

Thermostat Housing Design

Week 4 Exercise

Design Thermostat Enclosure Constraints

1. List all interface requirements (display, sensors, mounting)
2. Define the design envelope
3. Specify thermal requirements
4. Consider manufacturing constraints
5. Sketch initial housing concepts

Week 4 Summary

• CAD is essential for hardware prototyping
• Generative design accelerates optimization
• 3D printing enables rapid iteration
• Design guidelines ensure printable parts
• Services extend prototyping capabilities

Next week: Software Prototyping with AI

AI Code Generation Tools

Embedded Software Architecture

Thermostat Software Components

PID Control for HVAC

AI can help generate, optimize, and tune control algorithms

Week 5 Exercise

Prototype HVAC Control Algorithm

1. Define control requirements (setpoint, deadband)
2. Use AI to generate initial PID implementation
3. Create simulation environment for testing
4. Tune parameters for acceptable response
5. Document algorithm behavior

Week 5 Summary

• AI accelerates embedded software development
• Layered architecture enables parallel development
• Control algorithms can be AI-generated and refined
• API integration extends product capabilities
• Testing throughout catches issues early

Next week: Hardware/Software Co-Design

What is HW/SW Co-Design?

Simultaneous development of hardware and software with:

• Shared requirements understanding
• Interface definition at project start
• Parallel development streams
• Continuous integration

Interface Control Document (ICD)

Thermostat Sensor Interfaces

Power Budget

Week 6 Exercise

Define Interface Control Document

1. Document all sensor interfaces (electrical, software)
2. Define display interface specifications
3. Specify power architecture and budget
4. Define mechanical-electrical interfaces
5. Create interface diagram

Week 6 Summary

• Co-design enables parallel development
• Interface specifications must be defined early
• ICD documents interfaces for all teams
• Power budgets drive architecture decisions
• Clear interfaces reduce integration risk

Next week: Digital Twin Fundamentals

What is a Digital Twin?

A virtual representation of a physical system that:

• Mirrors physical system behavior
• Updates with real-world data
• Enables simulation and prediction
• Supports design and operation decisions

Digital Twin Maturity Levels

1. Digital Model: Static representation
2. Digital Shadow: One-way data flow (physical → digital)
3. Digital Twin: Bidirectional data flow
4. Predictive Twin: AI-enhanced prediction

Benefits for Prototyping

• Test before building physical prototypes
• Explore design alternatives virtually
• Reduce physical iteration cycles
• Validate control algorithms
• Train ML models with synthetic data

Thermostat Digital Twin

Week 7 Exercise

Model Thermostat Thermal Behavior

1. Define room thermal parameters (volume, R-value)
2. Model HVAC as on/off with capacity
3. Implement PID controller from Week 5
4. Simulate 24-hour cycle
5. Analyze control performance

Week 7 Summary

• Digital twins enable virtual testing
• Thermal modeling critical for thermostat accuracy
• Control algorithms validated in simulation
• AI can assist in model creation
• Reduces physical prototype iterations

Next week: Midterm Presentations

Midterm Expectations

By midterm, teams should have:

• Concept definition complete
• Initial requirements documented
• Interface specifications drafted
• First prototype iteration complete
• Digital twin model started

Presentation Requirements

• 10-15 minute presentation per team
• Demonstrate progress on thermostat prototype
• Show digital twin simulation results
• Discuss challenges and lessons learned
• Present plans for second half

Verification vs. Validation

• Verification: "Did we build the product right?"
  • Confirms implementation matches design
• Validation: "Did we build the right product?"
  • Confirms product meets user needs

Types of Simulation

Thermostat Validation Tests

Week 9 Exercise

Simulate Thermostat Control Response

1. Define test scenarios (heat wave, cold snap, daily cycle)
2. Configure simulation parameters
3. Run simulations and record results
4. Compare against requirements
5. Document failures and root causes

Week 9 Summary

• Simulation validates design before physical testing
• Multiple simulation types serve different purposes
• HIL testing bridges simulation and physical
• AI can generate comprehensive test cases
• Results guide design refinement

Next week: Physical Prototype Fabrication

From Virtual to Physical

Before physical fabrication:

• Design reviewed and frozen
• Manufacturing files generated
• Components sourced
• Assembly sequence planned
• Test procedures ready

PCB Ordering Options

FDM Print Settings

Assembly Sequence

1. Prepare housing (inserts, finish)
2. Assemble PCB (if not PCBA service)
3. Program and test bare PCB
4. Mount PCB in housing
5. Connect sensors and display
6. Close housing
7. System test

Week 10 Exercise

Plan Thermostat Prototype Build

1. Generate PCB manufacturing files
2. Order PCBs and components
3. Generate 3D print files
4. Print housing prototype
5. Plan assembly sequence with checkpoints

Week 10 Summary

• Physical fabrication requires careful planning
• PCB services enable rapid turnaround
• 3D printing settings affect performance
• Assembly sequence impacts quality
• AI can assist in manufacturing planning

Next week: Iterative Refinement

Fail-Fast Philosophy

• Seek failures early, not late
• Each failure teaches something
• Rapid cycles beat slow perfection
• Document every learning

Iteration Cycle

Metrics: Cycle time, Learning rate, Convergence

Root Cause Analysis

1. Observe the symptom
2. Gather data
3. Hypothesize causes
4. Test hypotheses
5. Identify root cause
6. Implement fix
7. Verify fix

Common Thermostat Issues

Week 11 Exercise

Design Iteration Experiment

1. Identify a performance gap in prototype
2. Hypothesize root cause
3. Design experiment to test hypothesis
4. Execute and collect data
5. Analyze results and implement fix
6. Verify fix effectiveness

Week 11 Summary

• Fail-fast accelerates learning
• Root cause analysis prevents recurring issues
• Structured experiments yield better results
• AI can assist in diagnosis
• Track convergence to know when to stop

Next week: Integration Testing

Integration Testing Goals

• Verify subsystems work together
• Find interface issues
• Validate system-level requirements
• Demonstrate functionality

Integration Strategy

Thermostat Integration Sequence

1. MCU + Power supply
2. + Display
3. + WiFi connectivity
4. + Temperature sensor
5. + Other sensors
6. + HVAC relay
7. Complete system

Week 12 Exercise

Create Thermostat Test Plan

1. Define integration sequence
2. Create test cases for each step
3. Define system-level test cases
4. Specify pass/fail criteria
5. Plan test documentation

Week 12 Summary

• Integration testing finds interface issues
• Incremental integration reduces risk
• Test automation improves efficiency
• Issue tracking ensures nothing is forgotten
• Validation confirms product meets needs

Next week: Prototype Documentation

Technical Data Package (TDP)

A TDP contains all technical information needed to:

• Understand the design
• Manufacture the product
• Test and verify compliance
• Maintain and support

TDP Contents

AI for Documentation

AI can help:

• Generate initial drafts
• Create consistent formatting
• Extract information from designs
• Generate drawing annotations

Important: AI output must be reviewed for accuracy!

Week 13 Exercise

Draft Thermostat TDP Sections

1. System specification (requirements summary)
2. Mechanical BOM with sources
3. Electrical schematic documentation
4. Software architecture document
5. Test summary report

Week 13 Summary

• TDP captures all design information
• Documentation enables reproducibility
• AI accelerates documentation creation
• Quality review remains essential
• BOM management critical for manufacturing

Next week: Prototype to Production Transition

Design for Manufacturing (DFM)

Design to:

• Minimize part count
• Use standard components
• Enable efficient assembly
• Reduce manufacturing variation
• Lower production cost

Design for Assembly (DFA)

• Minimize part count
• Design for top-down assembly
• Self-locating features
• Mistake-proofing (poka-yoke)
• Minimize fasteners

Prototype vs. Production Cost

Transition Checklist

• ☐ Design freeze complete
• ☐ All testing passed
• ☐ Documentation complete
• ☐ Manufacturing files verified
• ☐ Supply chain confirmed
• ☐ Quality plan defined
• ☐ Production test defined

Week 14 Exercise

DFM Analysis

1. Evaluate housing for injection molding
2. Review PCB for assembly automation
3. Identify cost reduction opportunities
4. Create production cost estimate
5. Document transition recommendations

Week 14 Summary

• DFM/DFA optimize designs for production
• Production costs differ from prototype costs
• Transition requires comprehensive documentation
• Lessons learned capture valuable knowledge
• Manufacturing feasibility must be validated

Final Presentation (15-20 min)

1. Product introduction and market context
2. System architecture and design decisions
3. Prototype demonstration
4. AI tools utilized and acceleration achieved
5. Testing results and validation
6. Manufacturing feasibility analysis
7. Lessons learned and recommendations

Final Report Sections

1. Executive Summary
2. Product Requirements
3. System Design and Architecture
4. Subsystem Designs
5. Digital Twin and Simulation Results
6. Prototype Development and Testing
7. Manufacturing Feasibility
8. Lessons Learned
9. Appendices (TDP)

Course Summary

Key Learning Outcomes:

1. Design and execute AI-accelerated prototyping workflows
2. Create functional prototypes in compressed timelines
3. Apply hardware/software co-design principles
4. Develop and utilize digital twins
5. Implement fail-fast methodologies
6. Evaluate manufacturing feasibility
7. Document prototype evolution

Course Materials

Start with: project_guide/00_project_overview.md

Thank You

Contact Information

Dr. David S. Cochran - cochrand@pfw.edu

Mr. Joseph Smith - smitjj09@pfw.edu

Office Hours: Wednesdays 2:30-4:30 PM, ETCS 229B

SE 54200 AI System Engineering Rapid Prototyping

Purdue University Fort Wayne