Project 4: AI Assistant – Class 7 Vocational Education Chapter 4 Notes & NCERT PDF

Class 7Vocational EducationChapter 4NCERT book: Kaushal Bodh - Vocational Education Class 7

Notes and a simple summary of Chapter 4 of the NCERT Class 7 Vocational Education book, with the official chapter PDF, flashcards, an MCQ quiz and an AI tutor for your doubts.

🎧 Listen to the content

Project 4: AI Assistant (Part 2: Work with Machines and Materials)

About the Project

Project 4: AI Assistant is the second example project in Part 2 – Work with Machines and Materials. You will learn about Artificial Intelligence (AI) and create an AI Assistant to help someone who has recently moved into your locality.

Intelligence is the ability to learn and to use what you have learnt in new situations. When machines become able to do this, it is called Artificial Intelligence (AI).

The book compares AI to how you learn names on the first day of school: after meeting people many times, your brain recognises them. Similarly, to teach a machine to recognise a banyan tree, we upload many different images — in shade and light, at different growth stages, from different angles and places — and associate them with the name 'banyan tree'. This process is called Machine Learning, and AI makes it possible.

AI can automate repetitive tasks, reducing human effort and increasing productivity (navigation apps, image recognition, translation, robots helping doctors or teaching).

But AI cannot feel emotions as humans do — it can only say an artwork "makes you feel good" if trained to do so.

What You Will Learn

Understand how AI can be a useful tool for humans.

Collect data to create an AI Assistant.

Train a machine to recognise data based on your instructions.

Create an AI Assistant to help someone learn about your locality.

Materials & Tools

Devices: a computer with webcam and microphone/speaker, or a tablet with Internet; a camera to take photos and record audio (or a parent's/teacher's smartphone).

AI tools named in the book: Teachable Machine (Google), Scratch, MIT AI Raise Playground, Quick Draw, Google Maps, Google Lens, Photomath, Bhashini (translation), GenerateStory.io, AutoDraw.

Step-by-Step Activities

Activity 1: Human vs Machine – who is better at what?

Speed Test: multiply 5643724 × 2 on paper, time yourself, then use a calculator. (Answer: 11287448 — every calculator gives the same answer, faster.)

Creative Guessing Game: draw 5 simple things in Quick Draw and count how often the computer guesses right.

Prediction Test: guess how long you'll take to reach school next Monday, then check Google Maps.

Conclusion to discuss: machines are faster at computing and guess accurately when they have enough information, but may fail when they need more.

Activity 2: AI can see, listen and speak

Listen: modern hearing aids use AI to suppress background noise and amplify speech, and can tell different speakers apart.

See: AI-enabled glasses with sensors help people with visual impairment.

Speak: AI uses recordings of people who lost their voice to turn whispers or stuttered speech into clear speech.

Prof. Stephen Hawking, paralysed by an illness from age 21, communicated using a system that selected words by tracking his cheek muscle movements and converted them to speech.

Try it yourself: Google Lens (identify a plant/object), Photomath (solve a maths problem step by step), Bhashini (translate "The school is close to my home" into another language).

Activity 3: Is AI creative?

Write your own story about a mela, festival or trip, then give AI a prompt to write one. An AI prompt is more detailed than search words — a set of instructions or questions that help AI give a specific reply. Improve the prompt with more details and compare. Use AutoDraw to clean up a rough sketch. Lesson: AI will never write a story like you do, but AI and you can help each other.

Activity 4: Preparing to design your AI Assistant

Decide what a newcomer should know: popular food, plants, animals, music, dance forms, places of tourist interest, communication (e.g., greetings Juley in Ladakh, Johar in Jharkhand). Data is any information we can see, hear or record — words, numbers, pictures, sounds, videos. This project focuses on image data.

Aisha's example: her mother is a fruit seller. Aisha trains a machine to sort mango varieties. She collects 12–15 images of each variety with different angles, backgrounds, brightness and sizes. Check each image: is it clear? Is there a shadow? Reject poor ones.

Activity 5: Teaching the machine to recognise images

Organise images into folders, one for each category/sub-category (e.g., plants → vegetable, flowering, fruit).

Activity 6: Training for recognition (Teachable Machine)

Open Teachable Machine → Image Project → Standard Image Model, and upload images into their classes.

Label each sub-category (e.g., Mango_Alphonso).

Click Train Model.

Activity 7: Testing and improving

Test in two ways: upload new images, or hold a real object in front of the webcam (real-time testing). If it makes mistakes, add more images to that category and retrain. Save with Export Model → Download Model.

Activity 8: Making it interactive

Load your model into MIT AI Raise Playground (Scratch-based) via File → Load from your Computer. Choose a character (friendly guide or historical figure), plan its questions, and give it a personality.

Activity 9: Sharing with others

Let people use your assistant, collect feedback, and note one change you'd make.

Extension: Teachable Machine can also train on audio — record more than ten clips per category (languages, bird sounds), train, test and retrain.

Safety First

Internet safety: make a list of dos and don'ts with your teacher and follow it; ask when in doubt; limit screen time — go out and play.

Sensitivity: get permission before taking photos (e.g., ask a shopkeeper); respect unfamiliar customs; don't harm plants and animals while photographing them.

Key Terms

Intelligence – ability to learn and use it in new situations (बुद्धि)

Artificial Intelligence (AI) – intelligence shown by machines (कृत्रिम बुद्धिमत्ता)

Machine Learning – teaching a machine by showing labelled data (मशीन लर्निंग / यंत्र अधिगम)

Data – information we can see, hear or record (आँकड़े / डेटा)

Image data – pictures used to train AI (चित्र डेटा)

Category / Sub-category – main group and smaller group (श्रेणी / उप-श्रेणी)

Label – name given to a class of data (लेबल / नाम-पट्टी)

Train model – making the machine learn from data (मॉडल प्रशिक्षण)

Testing – checking if the model recognises new data (परीक्षण)

Real-time testing – testing live through a webcam (वास्तविक-समय परीक्षण)

Prompt – detailed instruction given to an AI tool (निर्देश / प्रॉम्प्ट)

Algorithm – step-by-step instructions for getting work done by a machine (कलन-विधि / एल्गोरिदम)

Automate – make a machine do repetitive tasks (स्वचालित करना)

Screen time – time spent on devices (स्क्रीन समय)

Think and Answer – Hints

Enjoyed / challenges / do differently: e.g., the model confused two similar items → collect more varied images and retrain.

Flowchart (algorithm): 1. What do I want to do? → 2. Create the classification tree (categories, sub-categories) → 3. Collect and organise data → 4. Train the model → 5. Test → 6. Share. Writing these steps systematically is what programmers call an algorithm.

Jobs: data scientist, machine learning assistant, software engineer, robotics engineer, research scientist.

Career & Community Connections

Your AI Assistant helps newcomers to your locality. Talk to community elders, recent movers, parents and teachers to decide what matters. Careers linked to this work include data scientist, ML assistant, software engineer, robotics engineer and research scientist.

Common Mistakes

Uploading too few or too similar images — the machine can't recognise new ones.

Using unclear or shadowed photos without checking them.

Forgetting to label sub-categories clearly before training.

Not retraining after adding images to a weak category.

Forgetting to save/export the model.

Thinking AI feels emotions or is more creative than humans — it only does what it has been trained on.

Taking photos of people or shops without permission.

💡 Key Learning Points

✓AI is intelligence shown by machines; machine learning teaches a machine by showing it many labelled examples.
✓Machines are faster at computing and accurate with enough data, but cannot feel emotions or create like humans.
✓Good training data needs many varied images (angles, backgrounds, brightness, sizes) organised into labelled categories.
✓Teachable Machine workflow: upload images into classes, label them, train, test with new images or webcam, retrain, export.
✓An AI Assistant for newcomers can be made interactive in MIT AI Raise Playground, following internet safety and permission rules.

👨‍🏫 Teaching Tips

→Begin with the 'first day of school' analogy: ask how students learned classmates' names, then connect to machine learning.
→Run the Speed Test and Quick Draw game live in class to spark the 'human vs machine' discussion.
→Demonstrate Teachable Machine with two easy classes (e.g., pen vs pencil) before students build their own.
→Assign each group a locality category (food, plants, animals, music, places) so data collection is shared.
→Deliberately train a model with too few images to show errors, then add data and retrain to show improvement.
→Discuss AI's limits (emotions, creativity) and ethics: permission before photos, privacy, screen time.
→Ask students to write their project steps as a flowchart to introduce the idea of an algorithm.

📋 Assessment Questions

Can the student explain machine learning using the banyan tree or mango example?
Did the student collect enough varied, clear images and organise them into labelled folders?
Can the student train, test and retrain a model in Teachable Machine?
Can the student name one thing AI can do and one thing it cannot do?
Did the student follow internet safety and ask permission while collecting data?

Ask the AI tutor about Class 7 Vocational Education Chapter 4

Get question answers, important questions, a quick MCQ quiz or a simple explanation of “Project 4: AI Assistant”. Free — sign in with Google.

How the AI tutor works →

👩‍🏫 Class 7 Vocational Education classes near you — online or at home

Prefer a real teacher? TalentJR teachers take live online classes anywhere in India and visit homes in selected cities. Tell us what you need and our team will call you with timings and fees.

Live online classesSmall groups or 1:1 with a teacher, from home, any city.
Teacher at homeA TalentJR teacher visits your home (selected cities).
At a centreClasses at a TalentJR partner centre near you, where available.

Keep learning

← Previous: Chapter 3Project 3: Tie and Dye Next: Chapter 5 →Project 5: Storytime with Puppets