Strategies for Integrating Generative AI in Engineering Education

Author

Dr. Andrew Katz

Welcome!

Note

Please take a moment to consider:

  1. What course would you most like to integrate AI tools into?
  2. What’s your biggest concern about using AI in your teaching?

Workshop Goals:

  • Identify dimensions of AI integration in engineering education
  • Analyze case studies using a structured framework
  • Select appropriate AI approaches for your specific courses
  • Begin developing an implementation plan
  • Access resources for continued development

Workshop Schedule

  • Part 1: Foundation (15 min)
    • Introduction and AI landscape
    • The AI Integration Taxonomy
  • Part 2: Exploring the Taxonomy (40 min)
    • Dimensions of AI integration
    • Case examples across engineering disciplines
    • Discussion of implementation approaches
  • Part 3: Application & Planning (35 min)
    • Case study analysis in small groups
    • Individual implementation planning
    • Next steps and resources

PART 1: FOUNDATION

The AI Landscape in Engineering Education

Generative AI Tools in Engineering

  • Large Language Models (ChatGPT, Claude)
  • Code Generation (GitHub Copilot)
  • Image Generation (DALL-E, Midjourney)
  • Speech Recognition (Whisper)
  • Multimodal Tools (GPT-4V)

Engineering Education Challenges

  • Technical domain knowledge
  • Visualization of complex concepts
  • Skill development vs. conceptual understanding
  • Balancing theory and application
  • Preparing for evolving professional practice

AI Tools Evolution

Students and professionals already using these tools

Opportunity to intentionally integrate rather than react

Why a Framework for Integration?

Ad hoc integration leads to:

  • Inconsistent student experiences
  • Missed pedagogical opportunities
  • Assessment misalignment
  • Potential equity issues
  • Unclear expectations

A structured framework provides:

  • Common language for discussing integration
  • Multiple dimensions for consideration
  • Intentional decision-making
  • Alignment with educational goals
  • Disciplinary adaptability

The AI Integration Taxonomy

Six dimensions to consider:

  1. Pedagogical Purpose
    Why integrate AI?

  2. Integration Depth
    How deeply embedded?

  3. Student Agency
    How much student control?

  1. Assessment Alignment
    How to evaluate learning?

  2. Technical Implementation
    What technical aspects matter?

  3. Ethical & Professional
    What broader implications?

This framework helps map the integration landscape and make intentional choices

Full Taxonomy Document

PART 2: EXPLORING THE TAXONOMY

1. Pedagogical Purpose Dimension

Five primary purposes:

  • Conceptual Understanding
    Explaining complex concepts, addressing misconceptions

  • Skill Development
    Bypassing technical hurdles for higher-order skills

  • Process Augmentation
    Enhancing workflows and methodologies

  • Content Creation
    Generating or transforming educational materials

  • Visualization
    Helping visualize complex phenomena

Engineering Example:

In thermodynamics, students use ChatGPT to:

  • Generate multiple explanations of entropy concepts
  • Connect microscopic and macroscopic views
  • Create conceptual comparisons between similar processes
  • Identify and address common misconceptions

Case Study: Thermodynamics with ChatGPT

Pedagogical Purpose Dimension

Purpose Dimension: Discussion

Note

Small Group Discussion (3 minutes)

  1. What aspects of your courses are most challenging for students to understand?
  2. Which pedagogical purpose seems most valuable for your context?
  3. How might AI tools address these specific challenges?

Consider:

  • Conceptually difficult topics
  • Areas where students struggle with visualization
  • Skills that require significant practice
  • Content that needs multiple perspectives

2. Integration Depth Dimension

Spectrum of integration:

  • Supplemental Resource
    Optional tools outside core instruction

  • Guided Integration
    Structured prompts for specific activities

  • Embedded Practice
    AI integrated throughout regular coursework

  • Transformative Redesign
    Course restructured around AI capabilities

Engineering Example:

In data structures course with GitHub Copilot:

  • Started as supplement for debugging
  • Moved to guided exercises comparing manual and AI-assisted implementation
  • Evolved to embedded practice with design-first approach
  • Assessment redesigned to focus on algorithm design over syntax

Case Study: Data Structures with GitHub Copilot

Integration Depth Spectrum

3. Student Agency Dimension

Levels of student choice and responsibility:

  • Instructor-Directed
    Faculty provides specific prompts/tools

  • Scaffolded Autonomy
    Progressive responsibility with guidance

  • Guided Exploration
    Students experiment within boundaries

  • Full Autonomy
    Independent decisions about AI use

Engineering Example:

In a materials science course:

  • Began with specific instructor-provided prompts
  • Gradually introduced template libraries students could modify
  • Moved to student-created prompts with guidance
  • Culminated in students determining when/how to use AI tools

Case Study: Materials Science with Claude

Student Agency Levels

Integration & Agency: Quick Poll

Note

With a partner, take two minutes to discuss:

  1. Current integration depth in your courses
  2. Desired integration depth you’d like to achieve
  3. Student agency level you’d be comfortable with

Questions to consider:

  • What barriers exist to deeper integration?
  • What student preparation would be necessary?
  • How might agency levels progress across a program?

4. Assessment Alignment Dimension

Assessment approaches:

  • Process Documentation
    Evaluating AI use in workflow

  • Comparative Analysis
    Evaluating AI outputs vs. alternatives

  • Critical Evaluation
    Verifying and refining AI contributions

  • Meta-Learning
    Reflection on learning with AI

  • AI-Restricted Components
    Some assessment without AI

Engineering Example:

In electrical engineering circuit design:

  • Students document their prompting strategies
  • Compare AI suggestions with manual calculations
  • Identify and correct errors in AI recommendations
  • Reflect on how AI affected their design process
  • Still complete fundamental circuit analysis manually

Case Study: Electrical Engineering with Claude

Assessment Alignment

5. Technical Implementation Dimension

Implementation aspects:

  • Tool Selection
    Matching capabilities to objectives

  • Access Provision
    Ensuring equitable student access

  • Prompt Engineering
    Developing effective prompts

  • Error Management
    Handling AI limitations

  • Integration Infrastructure
    Technical platforms for delivery

Engineering Examples:

  • Civil engineering using Whisper for field note transcription
  • Chemical engineering using DALL-E for safety visualization
  • Mechanical engineering using ChatGPT for ideation
  • Each tool selected for specific capabilities aligned with learning goals

Case Studies: Tool-Specific Applications

Technical Implementation

Assessment & Technical: Think-Pair-Share

Note

Think-Pair-Share (5 minutes)

  1. What assessment challenges do you anticipate with AI integration?
  2. What technical implementation concerns are most relevant in your context?
  3. What strategies might address these challenges?

Consider:

  • Balancing individual mastery with AI assistance
  • Technical resources available in your department
  • Student access and equity issues
  • Discipline-specific technical needs

6. Ethical & Professional Development Dimension

Ethical aspects:

  • Attribution Practices
    Citation of AI contributions

  • Professional Norms
    Alignment with industry practices

  • Critical AI Literacy
    Understanding capabilities and limitations

  • Responsible Use
    Ethical decision-making

  • Equity Considerations
    Benefits reaching all students

Engineering Example:

Across disciplines:

  • Developing AI contribution statements
  • Consulting with industry on current practices
  • Teaching systematic verification of AI outputs
  • Discussing societal implications of AI in engineering
  • Addressing varying levels of prior AI experience

Cross-Cutting Ethical Considerations

Ethical Considerations

Ethical & Professional: Brief Discussion

Important

Brief Discussion

What ethical considerations are particularly important in your engineering discipline?

Consider:

  • How is your industry using AI tools?
  • What ethical concerns are specific to your field?
  • How might AI use affect different student populations?
  • What professional skills should students develop?

Implementation Examples by Discipline

We’ve created detailed example implementations that map to different positions on the taxonomy:

These examples include detailed implementation steps, sample prompts, and assessment strategies

Case Study Analysis

Prompt Engineering in Engineering Education

Key principles for effective prompts:

  • Be specific about engineering context
  • Include relevant technical parameters
  • Clarify expected detail/technical level
  • Request verification steps
  • Include format specifications
  • Consider iterative prompt chains

Example: From general to specific
❌ “Explain entropy”
✅ “Explain entropy from statistical mechanics perspective for junior-level thermodynamics students”

Sample prompt structures:

  1. Conceptual explanation prompt:
    • Multiple perspective requests
    • Connection to applications
    • Misconception identification
  2. Technical verification prompt:
    • Solution review with error identification
    • Underlying principle explanation
    • Alternative approach suggestion
  3. Design exploration prompt:
    • Multiple solution generation
    • Constraint-based evaluation
    • Trade-off analysis

See examples directory for discipline-specific prompt templates

Generative Tools

Assessment Redesign Principles

Moving beyond traditional assessment in AI-integrated courses:

  • Assess the Process: Documentation of AI interactions, decision-making, verification
  • Focus on Higher-Order Skills: Engineering judgment, critical evaluation, constraint analysis
  • Maintain Knowledge Verification: Targeted components without AI assistance
  • Balance Product and Process: Evaluate both outcomes and the methods used
  • Clear Attribution Standards: Consistent guidelines for documenting AI contributions
Tip

Rubric elements should reward critical thinking about AI outputs, not just the final product quality. Good rubrics include evaluation of verification strategies and decision rationale.

See our comprehensive Assessment Redesign Guide for detailed rubrics and examples

Implementation Challenges & Solutions

Common Challenges:

  • Varying student AI literacy levels
  • Technical accuracy verification
  • Equity of access to AI tools
  • Balance of efficiency vs. learning
  • Academic integrity considerations
  • Rapid tool evolution
  • Student overreliance on AI

Effective Solutions:

  • Scaffolded introduction with baseline training
  • Create verification protocols and checklists
  • Provide institutional or classroom access
  • Focus assessment on process and reflection
  • Develop clear attribution guidelines
  • Emphasize transferable skills beyond tools
  • Design assignments requiring critical evaluation

Challenge intensity varies by integration depth and student agency level

Integration Challenges

PART 3: APPLICATION & PLANNING

Case Study Analysis Activity

Small Group Activity (15 minutes):

  1. Each group will analyze one case study using the taxonomy
  2. Use the Case Study Analysis Worksheet provided
  3. Map the case to each dimension of the taxonomy
  4. Identify key integration decisions and their rationale
  5. Discuss how similar approaches might apply to your contexts
  6. Select one key insight to share with the full group

Case Studies Available

Tip

Each case study maps to different positions along the taxonomy dimensions

Implementation Planning

Individual Work (15 minutes):

  1. Select one course for potential AI integration
  2. Complete the Implementation Planning Template
  3. Map current position on each taxonomy dimension
  4. Choose target positions that align with course goals
  5. Identify specific implementation steps for 1-2 priority dimensions
  6. Document anticipated challenges and resources needed
  7. Share plan with a partner for feedback (5 minutes)

Implementation Timeline

Example Implementation Plan

Course: Thermodynamics II (Junior-level)

Current State: * No formal AI integration * Students using AI unofficially * Traditional problem-based assessment * Conceptual understanding challenges with entropy, availability, and multi-scale phenomena

Priority Dimensions: 1. Pedagogical Purpose (Conceptual Understanding) 2. Assessment Alignment (Process Documentation) 3. Student Agency (Scaffolded Autonomy)

Implementation Actions: 1. Create prompt library for thermodynamic concepts 2. Develop visual representation activities using image AI 3. Design concept mapping assignment with AI feedback 4. Implement verification protocols for AI explanations 5. Create process portfolio assessment structure 6. Pilot with entropy unit before full implementation

Resources Needed: * Example prompt collection for key concepts * Image generation tool access (DALL-E) * LMS integration for documentation * Assessment rubrics focused on concept mastery * Student guidance for critical AI evaluation

Based on the detailed example in our workshop materials

Key Takeaways

  • Intentional integration is key to effective AI use in engineering education

  • The taxonomy framework provides multiple dimensions to consider for implementation

  • Different dimensions may be prioritized based on specific course challenges and goals

  • Progressive implementation often works better than complete redesign

  • Assessment alignment is critical for meaningful integration

  • Consider implications for student professional development

Resources Available

Contact Information: Dr. Andrew Katz Email: [email protected]

Next Steps & Questions

Next Steps: 1. Complete your implementation plan 2. Identify one small action to take in the next month 3. Consider forming discipline-specific implementation groups 4. Explore additional resources provided

Important

Questions & Discussion
What questions do you have about implementing AI in your engineering courses?

Thank you for your participation!