CBSE Class 10 · Artificial Intelligence, code 417 · Project ideas 2026-27

Class 10 AI project ideas that are small enough to finish and real enough to defend.

These class 10 AI project ideas are for the 10-mark project in CBSE 417 Part C and the 5-mark viva that follows it. There are fifteen, graded easy, medium and stretch. Each is tied to one Sustainable Development Goal and one AI domain, Statistical Data, Computer Vision or NLP, and each comes with its data, model and metric for every stage of the project cycle. In our live batch, students pick one in the first term and build it with a teacher checking each stage.

Book a free demo class Book a priority demo Ask on WhatsApp

Live online · 15 original project ideas below · each with its SDG and AI domain · built in Orange or Python · project viva rehearsed · first class free

Start here

Courses that help a project get built

The 417 course whose batch guides the project from idea to viva, the Orange course for building the model without code, and the data science course for students who want their project in Python.

CBSE Computational Thinking and AI for Teens course thumbnail

The batch / 417 Class 10

CBSE AI for Class 9 to 12

The full 417 course, with the project chosen in the first term and each stage reviewed in the Wednesday lab.

Open the course →
Visual Machine Learning with Orange for Teens course thumbnail

Easy and medium / Orange

Visual ML with Orange

Image, text and table workflows in Orange, the no-code route that most of the ideas below can be built in.

Open the course →
Data Science for Teens course thumbnail

Stretch / Python

Data Science for Teens

Pandas, charts and simple models in Python, for students taking one of the stretch ideas in code.

Open the course →

Ans. The short version

Good class 10 AI project ideas for CBSE 417 are small, linked to one Sustainable Development Goal and built in one AI domain: Statistical Data, Computer Vision or NLP. Easy options include a leaf disease checker, a fruit ripeness sorter or a sentiment analysis project on clinic reviews; harder ones include coral bleaching detection or a complaint classifier. Each should follow the AI project cycle and report real evaluation numbers. The project carries 10 marks and its viva 5. Modern Age Coders guides projects from idea to viva in a live Monday and Wednesday 9:00 PM IST batch, or one to one.

Q1. What makes a good project?

What makes a good AI project for class 10 SDG work

A good Class 10 AI project answers one clear question about one real problem, uses data you can get, and ends with numbers you can explain. CBSE asks that it relate to the Sustainable Development Goals, and its own sample, penguin species prediction, shows how small the scale can be.

Grade on this pageDataToolTypical build time
EasyA ready open dataset, or a few dozen photos you takeOrange, no codeThree to four weekends
MediumAn open dataset that needs cleaning, or a survey you runOrange, or short PythonFive to six weekends
StretchData you collect and label yourselfPython, comparing two modelsEight weekends or more

Statistical Data

Rows and columns of numbers or categories: readings, surveys, records. The model predicts a label or a value from the columns.

Computer Vision

Images sorted into labelled folders. The model learns to tell the classes apart from pixel patterns.

Natural Language Processing

Short texts such as reviews, headlines or messages. The model scores or sorts them by their words.

Build times are from our batch's experience, not a CBSE rule. Easy does not mean fewer marks: a finished, honest easy project beats an abandoned ambitious one.

Q2. Fifteen ideas, easy to stretch

Easy AI project class 10 ideas, then medium and stretch

Each idea lists its goal, domain and a one-line run through the cycle: scope, data, explore, model, evaluate. All fifteen are written for this page; adapt them to your town and your data.

Easy: build in Orange with data you can get this week

  1. Leaf health checker. SDG 2, Zero Hunger · Computer Vision.
    Scope: small farmers notice leaf disease late. Data: your own photos of healthy and spotted leaves from one crop. Explore: view samples in Image Viewer. Model: Image Embedding with Logistic Regression. Evaluate: recall on diseased leaves, since a missed one spreads.
  2. Fruit ripeness sorter. SDG 12, Responsible Consumption · Computer Vision.
    Scope: shops throw away fruit that ripened unnoticed. Data: photos of bananas at three stages, taken over a week. Explore: are lighting and background consistent? Model: an image classifier in Orange. Evaluate: confusion matrix; which stages get mixed up.
  3. Sentiment analysis of clinic reviews. SDG 3, Good Health · NLP.
    Scope: what patients praise or complain about at local clinics. Data: public reviews, with names removed. Explore: Word Cloud after preprocessing. Model: Orange's Sentiment Analysis widget. Evaluate: label 50 reviews yourself and count agreement; 41 of 50 would be 0.82.
  4. Air quality band predictor. SDG 11, Sustainable Cities · Statistical Data.
    Scope: residents cannot read raw pollutant numbers. Data: an open city air-quality dataset. Explore: distributions of PM2.5 by band. Model: a classifier predicting the band from readings. Evaluate: accuracy, then which neighbouring bands it confuses.
  5. Sleep and screen-time survey. SDG 3, Good Health · Statistical Data.
    Scope: do students who use screens late feel less rested? Data: an anonymous survey of classmates, with consent. Explore: a scatter plot of screen time against sleep hours. Model: a classifier for "rested" or "not rested". Evaluate: accuracy, and an honest note on the small sample.

Medium: more cleaning, a harder question or some Python

  1. Coral bleaching detection. SDG 14, Life Below Water · Computer Vision.
    Scope: reef surveys need to flag bleached coral quickly. Data: open image sets of healthy and bleached coral on data-sharing sites. Explore: check the two classes are balanced. Model: Image Embedding with a classifier. Evaluate: recall on bleached coral (worked below).
  2. Misleading headline flagger. SDG 16, Peace and Justice · NLP.
    Scope: false headlines spread faster than corrections. Data: an open labelled headline dataset. Explore: frequent words in each class. Model: Bag of Words with a classifier. Evaluate: precision, because calling a true story false does harm.
  3. Hand sign alphabet reader. SDG 10, Reduced Inequalities · Computer Vision.
    Scope: few people can read sign language. Data: your own photos of five letter signs from several hands. Explore: are signs from different hands alike? Model: an image classifier. Evaluate: per-letter accuracy, and which letters look alike.
  4. Forest fire risk from weather. SDG 15, Life on Land · Statistical Data.
    Scope: forest staff need warning of high-risk days. Data: an open dataset of weather readings with fire or no-fire labels. Explore: temperature and humidity on fire days. Model: a decision tree. Evaluate: recall on fire days.
  5. Household electricity use. SDG 7, Clean Energy · Statistical Data.
    Scope: families do not know which habits raise their bills. Data: a survey of appliances and monthly units from twenty or more homes. Explore: units against AC hours. Model: classify "high" or "normal" use. Evaluate: accuracy, with the limits of a small sample.

Stretch: your own labelled data, two models compared

  1. Civic complaint sorter. SDG 11, Sustainable Cities · NLP.
    Scope: complaints reach the wrong department. Data: 150 or more short complaints you write or collect and label as water, roads or power. Explore: top words per label. Model: Bag of Words, two classifiers compared. Evaluate: confusion matrix across three classes.
  2. Pothole spotter. SDG 9, Infrastructure · Computer Vision.
    Scope: potholes are reported late. Data: your own road photos, labelled pothole or clear, taken safely from footpaths. Explore: variety of light and surface. Model: two classifiers compared. Evaluate: recall on potholes, then false alarms.
  3. Camera-trap filter. SDG 15, Life on Land · Computer Vision.
    Scope: volunteers sift thousands of empty trap frames. Data: open camera-trap images labelled animal or empty. Explore: night versus day frames. Model: image classifier in Python or Orange. Evaluate: recall on animal frames, so none are discarded.
  4. Solar output estimator. SDG 7, Clean Energy · Statistical Data.
    Scope: rooftop owners want to know what a cloudy week costs them in power. Data: open solar generation and weather records. Explore: output against cloud cover. Model: regression in Python or Orange. Evaluate: average error in units, compared with a simple guess.
  5. Mixed-language climate opinions. SDG 13, Climate Action · NLP.
    Scope: many posts mix Hindi and English, which English sentiment tools misread. Data: 200 or more posts you label positive, negative or neutral. Explore: words the English tool gets wrong. Model: your own Bag of Words classifier against the ready widget. Evaluate: which wins, and on what.

Q3. Which idea fits you?

Choosing your idea, and a coral bleaching project worked through

Pick the idea whose data you can actually get and whose question you care about; that interest is what carries you through the viva.

If you areStart withWhy
New to Python and short of timeIdeas 1, 2 or 3All Orange, data ready in days
Comfortable with tables and chartsIdeas 4, 9 or 10Statistical Data suits the Unit 4 practical too
Interested in imagesIdeas 6, 8 or 12Strong viva talking points on pixels and recall
Interested in languageIdeas 3, 7 or 11Directly uses the Unit 6 theory
Confident in Python, aiming highIdeas 13, 14 or 15Own data and model comparison show depth

Idea 6: the 4Ws canvas, filled

  • Who: reef survey teams and the fishing communities that depend on reefs.
  • What: bleached coral is spotted late, after many survey photos pile up unchecked.
  • Where: in survey photos from shallow tropical reefs.
  • Why: earlier flags let protection efforts start sooner, serving SDG 14.
Evaluating idea 6bleached = positive
Test set: 60 bleached, 60 healthy images.

              Actual: bleached  Actual: healthy
Pred bleached     TP = 51          FP = 6
Pred healthy      FN = 9           TN = 54

Accuracy  = 105 / 120 = 0.875
Precision = 51 / 57   = 0.89
Recall    = 51 / 60   = 0.85

An illustrative result, not a real model's. The sentence that earns viva marks: "Nine bleached corals were missed, so recall matters most; I would add more bleached images."

How to write each stage up in the report is on the project and practical file page, and the metrics above are drilled on evaluating models questions.

Q4. Which ideas cause trouble?

Project ideas to avoid, and the safer version of each

Some ideas look exciting but fail on data, ethics or time. Each has a smaller cousin that works.

Risky ideaThe problemSafer version
Face recognition attendance for your classFaces of minors collected without proper consent raise privacy issuesHand sign reader (idea 8) on your own hands
An app that diagnoses a diseaseClaims a medical result a student project cannot supportLeaf health checker (idea 1), or clinic review sentiment (idea 3)
Predicting the share marketNo clear SDG, and no way to judge success in a termHousehold electricity use (idea 10)
A chatbot that answers everythingToo broad to build or evaluateCivic complaint sorter (idea 11), one task done well
Any idea with no data sourceStalls at data acquisitionPick an idea whose data you can see before you commit

Ethics is part of the project viva, so mention consent, privacy and possible bias in your report. The vocabulary is revised on project cycle and ethics, and the questions you will be asked about your project are on class 10 AI viva questions.

Q5. How is a project guided online?

How the live batch takes a project from idea to viva

The teacher never builds the project. The student does, with checkpoints that stop small problems becoming late ones.

Idea and data check

Each student shortlists two ideas from a list like this one, then shows the class the data for each. The idea with data in hand wins.

Stage reviews on screen

One Wednesday a month, students share their Orange canvas or notebook, and the teacher asks what each step does and why.

Evaluation before polish

Numbers come before the report: the teacher checks the test split and confusion matrix, then the student writes up what the model gets wrong.

If your school asks for a Python project, the libraries are covered on Python for CBSE Class 10 AI; for the Orange route, see Orange Data Mining for Class 10 AI.

Batch or personalised: you choose

Build your project in the Monday and Wednesday batch, or one to one

Both formats follow the same checkpoints with the same teachers. The batch shows students fifteen other projects taking shape, which sharpens their own. One to one gives a project the full hour every time. You pay month by month with no admission fee, and can stop whenever a month ends.

Option 1 · Fixed batch timings

Group batch, Monday and Wednesday 9 PM

₹1,499

per month, two live classes a week

  • Live on video every Monday and Wednesday from 9:00 PM IST
  • A class of fifteen to twenty, all building their own project
  • Idea and data chosen in the first term
  • A monthly stage review on the shared screen
  • Project viva rehearsed before the school date
Book a free demo in this batch

Option 2 · Choose your own timing

Personalised one to one

₹4,999

per month, one private class a week, at a time set with you

  • Sessions booked at hours that suit your family
  • Every session spent on your child's own project
  • Help with data collection plans and consent forms
  • Python or Orange, whichever the school prefers
  • The teachers who run the 417 batch
Ask for a personalised plan

Students starting late, or taking a stretch idea in Python, often move faster one to one. Students still learning the units the project depends on are better placed in the batch, where theory and project move forward side by side. The fee appears in your local currency.

Studying outside India?

Class 10 AI project help for CBSE schools in the Gulf and abroad

CBSE schools abroad set the same Part C project, and local problems make strong SDG links: water in the Gulf, reefs off Oman and the UAE. The batch table below gives its local start for this exam winter.

Where you liveMon and Wed 9:00 PM ISTBest format
India (IST)
every Indian city		9:00 PMBatch fits
UAE and Oman
Dubai, Abu Dhabi, Sharjah, Muscat		7:30 PMBatch fits
Saudi Arabia, Qatar, Kuwait, Bahrain
Riyadh, Jeddah, Dammam, Doha, Kuwait City, Manama		6:30 PMBatch fits
Nepal
Kathmandu		9:15 PMBatch fits
Ethiopia and East Africa
Addis Ababa and East Africa Time		6:30 PMBatch fits
Singapore and Malaysia
Singapore, Kuala Lumpur		11:30 PMChoose 1-on-1

When the batch clashes with your evening, project sessions can be booked one to one at your hour. Families outside India will find timings and formats on our page for Gulf and overseas CBSE students.

What families say

Rated 4.9 across 547 Google reviews

Real reviews from real families. We neither write nor commission them.

★★★★★

"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."

Ria Mukherjee

Parent

★★★★★

"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."

Samriddha Mondal

Student

★★★★★

"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."

Vansh Agarwal

Student

★★★★★

"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."

Sonam Oswal

Parent of Dhairya

★★★★★

"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."

Sonu Goyal

Parent

★★★★★

"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."

Ritu Kedia

Parent

The rest of the Class 10 series

More pages for Class 10 board students

Every page in the CBSE Class 10 AI (417) series, grouped by what you need: classes, each unit, question practice, notes and revision, and the exam itself.

Classes, tuition and batches

Unit by unit

Questions and practice

Notes and revision

Exam, marks and practicals

Choosing the subject

ICSE Class 10 Computer Applications (Java): The Tue and Wed ICSE Java batch ICSE Computer Applications syllabus ICSE Java programs practice ICSE CA important questions

Questions about Class 10 AI projects

Class 10 AI project ideas: questions families ask us

What are some easy AI project ideas for class 10?

A leaf health checker from photos, a fruit ripeness sorter, a sentiment analysis of clinic reviews, an air quality band predictor or a survey on sleep and screen time. Each can be built in Orange without code, uses data you can get within a week and links to one Sustainable Development Goal.

Does the class 10 AI project have to be linked to an SDG?

CBSE asks that the Class 10 AI project relate to the Sustainable Development Goals. Name the goal in your problem scoping, explain who benefits, and keep the link specific: for example SDG 14, Life Below Water, for a coral bleaching project, rather than a general claim about helping the planet.

How do I do a sentiment analysis project in class 10?

Collect short texts such as public reviews, remove names, and load them into Orange with the Corpus widget. Clean them with Preprocess Text, score them with Sentiment Analysis and view the results in a Data Table. To evaluate, label about fifty texts yourself and count how often the tool agrees with you.

Is coral bleaching a good class 10 AI project?

Yes, as a medium-level Computer Vision project for SDG 14. Use an open image set of healthy and bleached coral, build an image classifier in Orange with Image Embedding, and report recall on bleached coral, since a missed bleached reef is the costly error. Check that both classes have similar numbers of images.

What is the class 10 AI project worth in marks?

The project carries 10 marks and the viva on it carries 5, out of the 50 practical marks in Part C of CBSE 417. CBSE allows a project, a field visit or a student portfolio for those 10 marks; your school tells you which it expects.

Can I make the class 10 AI project without coding?

Yes. Orange, which is part of the 417 curriculum, lets you build image, text and table models by connecting widgets. Every idea on this page can be built in Orange, including the stretch ones, though students confident in Python often find comparing models quicker in code.

Which AI domain should my project use?

One of the three in the curriculum: Statistical Data for tables of numbers, Computer Vision for images, or Natural Language Processing for text. Pick the domain whose data you can most easily get, and keep to one domain so the project stays finishable.

Can I use a dataset from the internet for my AI project?

Yes, open datasets are fine if you name the source in your report and can explain what each column or folder means. Data you collect yourself, with consent where people are involved, is often easier to defend in the viva because you know exactly how it was made.

How long does a class 10 AI project take?

An easy Orange project takes about three to four weekends of steady work; a stretch project with your own labelled data can take eight or more. Start in the first term, so the project is finished well before the school practical and the theory paper.

Do you help students with their AI project?

Yes. The project is guided inside the batch and one to one sessions: choosing an idea, checking the data, reviewing each stage and rehearsing the viva. The student builds the project themselves. There is no separate project charge: it sits inside the monthly fee on this page, the first class is a free demo and nobody pays an admission fee.

Start

Shortlist a project idea in a free class

During the free demo, your child talks through two or three ideas from this page with a teacher, checks which one has data they can actually get, and sketches its 4Ws canvas. You leave with a clear first step, and the mentor tells you whether the batch or one to one would suit the project better.

Revising the theory your project leans on? The units are summarised on Class 10 AI notes, unit wise.

WhatsApp us · +91 91233 66161 · connect@modernagecoders.com

Every class is live video. The form requests a callback and reserves nothing.

We call within a few hours. No card, no enrolment fee.

Want it sooner? Book a priority demo

A full live class of about 45 to 60 minutes at a slot you pick, with a written skill report and learning roadmap on WhatsApp afterwards. The fee is adjusted against your first month if you enrol.

Book a priority demo

Keep exploring Modern Age Coders

By class and age

Learn more

Free resources

From the blog

Start here