Introduction to AI and Prompt Engineering

Course objectives
1. Understand what Artificial Intelligence is, its history, and how it is changing every engineering field 2. Understand key AI concepts — Machine Learning, Deep Learning, and Neural Networks — at a conceptual level 3. Explore how AI is applied across domains — healthcare, agriculture, manufacturing, finance, and education 4. Understand Generative AI — how large language models (LLMs) like ChatGPT and Gemini work 5. Learn and practise Prompt Engineering — the skill of communicating effectively with AI tools to get useful results

Course content
Unit Topics Weightage Introduction to Artificial Intelligence 1. Definition and scope of AI; brief history — from 1. Alan Turing to ChatGPT 1. Types of AI — Narrow AI, General AI, Super AI (concept) 2. AI vs Machine Learning vs Deep Learning — what is the difference 3. How machines learn — data, patterns, and predictions (concept, no maths) 4. AI today — where it already exists in daily life (Google Maps, Spotify, Netflix, Google Photos, voice assistants) 5. Machine Learning Concepts, Learning Strategies- Supervised, Unsupervised, Reinforcement Learning 6. Data, Data Visualization, Data Pre-processing Class Activity: Students list 5 AI tools they use every day and discuss how each one 'learns'. Branch connect: CE — AI in structural health monitoring; ME — AI in predictive maintenance; EE — smart grid; CSE — AI in compilers; BT — drug discovery; AG — crop yield prediction. Core AI Concepts 1. Machine learning: Classification Techniques- K- Nearest Neighbors (KNN), Naïve Bayes, Clustering Technique-K-Means 2. Classification metrics-accuracy, precision, recall, F1 score, MAE. 3. Neural networks — concept of neurons, layers, weights (simple diagram) 4. Deep learning — what makes it 'deep'; why it needs GPUs, CNN 5. Computer vision — how AI sees images (face recognition, medical imaging) 6. Natural Language Processing (NLP) — how AI understands text and speech Class Activity: Show Google Teachable Machine (teachablemachine.withgoogle.com) — train a simple image classifier in 5 minutes, no code. Branch connect: ECE — image sensors + AI; BT — protein structure prediction (AlphaFold); AG — plant disease detection from photos; TX — fabric defect detection. Generative AI and Large Language Models 1. What is Generative AI — AI that creates text, images, code, audio, video 2. Large Language Models (LLMs) — what they are, how they are trained (concept) 3. Popular LLMs — ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic), Llama (Meta) 4. Tokens, context window, temperature — simple explanation 5. Multimodal AI — text + image + voice in one model (GPT-4o, Gemini concept) 6. AI image generation — Stable Diffusion, Midjourney, DALL-E (concept + demo) 7. Limitations of LLMs — hallucination, knowledge cutoff, bias 8. Responsible use — academic integrity, plagiarism, AI-generated content disclosure Class Activity: Students interact with ChatGPT/Gemini/Claude; observe how different inputs give different outputs; identify one hallucination example. Every professional today uses at least one Generative AI tool. Understanding how LLMs work makes students better users and future developers. AI Applications Across Engineering Domains 1. AI in Civil Engineering — structural health monitoring, smart cities, traffic optimisation 2. AI in Mechanical Engineering — predictive maintenance, design optimisation, robotics 3. AI in Electrical and Electronics — smart grids, fault detection, VLSI design automation 4. AI in CSE — code generation (GitHub Copilot), cybersecurity, natural language interfaces 5. AI in Biotechnology and Healthcare — drug discovery, medical imaging, genomics, wearables 6. AI in Agricultural Engineering — precision farming, crop monitoring, drone-based spraying, soil analysis 7. AI in Chemical and Textile Engineering — process optimisation, quality control, material design 8. Ethical concerns — bias in AI, data privacy, job displacement, responsible AI Class Activity: One case study per student group — each group presents a 2-minute real AI application from their own engineering branch. This unit ensures every student, regardless of branch, sees AI as directly relevant to their future career. Prompt Engineering 1. What is a prompt — the instruction you give to an AI model 2. Why prompt quality matters — same AI, very different results with different prompts 3. Principles of good prompting — clarity, context, specificity, role assignment, format instruction 4. Prompting techniques: • Zero-shot prompting — direct question • Few-shot prompting — give examples in the prompt • Chain-of-thought prompting — ask AI to think step by step • Role prompting — 'Act as a civil engineer and...' 5. Prompt for different tasks — summarisation, explanation, code generation, report writing, brainstorming, translation 6. Domain-specific prompting — students write prompts relevant to their own branch (CE, ME, BT, AG, etc.) 7. AI tools for engineers — GitHub Copilot (code), Grammarly (writing), Consensus (research), Wolfram Alpha, Canva AI, Notion AI Hands-on Practice (every lecture): Students write prompts, compare outputs, refine and improve. Graded prompt assignment: write 5 prompts for an engineering problem from their own branch. Prompt Engineering is already a job role (avg. salary $100K+ globally). Every engineer who can communicate well with AI has a significant career advantage. Suggested Assignments / Practice Tasks but not limited to Assignment Task Research task — no AI tool interaction required. Find 3 real AI applications in your engineering branch. For each: Assignment name the product/system, what data it uses, and what it 1 (Unit I) predicts or generates. Submit as a one-page note (250– 300 words). Concept-check task. Pick one real ML-based system relevant to your branch (e.g., a recommendation engine, Assignment a defect detector, a chatbot). Identify: (a) Is it supervised, 2 (Unit II) unsupervised, or reinforcement learning? (b) What could cause it to overfit or underfit? (c) Sketch a simple diagram showing its inputs, a basic neuron/layer concept, and output. Submit as a half-page note (200–250 words) with diagram. Hands-on task — direct AI interaction required. Interact with any free LLM (ChatGPT/Gemini/Claude). Ask it a Assignment question related to your branch. Identify one correct and 3 (Unit III) one incorrect (hallucinated) response. Explain why the hallucination occurred (150–200 words), with screenshots of both responses. Group case study: Each group (3–4 students) picks one AI application in their domain, prepares a 2-minute Assignment presentation covering: problem → data → AI solution → 4 (Unit IV) impact. Submit slides/notes plus a 1-page group summary handout. Write 5 prompts for an engineering problem from your branch using: (a) Zero-shot (b) Few-shot (c) Chain-of- Assignment thought (d) Role prompting (e) Format-specific prompt. 5 (Unit V) Submit the prompt + AI response + your evaluation of the output quality, as a single document (2–3 pages). Use an AI tool to help complete one academic task (summarise a topic, generate a study plan, explain a concept, translate a text). Document the prompt used, Assignment output received, and how you verified/corrected it. In a 6 (Unit V) 200–300 word reflection, address: (a) what you changed or verified and why, and (b) one way AI use in this task could raise a bias, privacy, or academic-integrity concern.