South Africa Coding Robotics Family AI Literacy: A Practical Home Guide

The Department of Basic Education (DBE) says Coding and Robotics for the Foundation Phase was endorsed and rolled out in 2025. It has also described AI literacy as increasingly essential for learners, teachers, and parents. For a parent, that does not mean buying a robot or giving a child unrestricted access to an AI tool. It means helping them understand instructions, patterns, systems, information, and responsible choices.

What the South African focus is trying to build

Coding and robotics are often mistaken for a screen subject. At their core, they develop habits of mind: breaking a large task into steps, testing a plan, spotting an error, improving a design, and explaining why a result happened. Robotics adds a physical system: something that senses, moves, follows instructions, or solves a practical problem.

For Foundation Phase children, this can begin without a computer. A child who gives clear directions to move a toy from the table to a box is practising sequencing. A child who changes those directions after the toy hits an obstacle is debugging. These are genuine coding ideas, even when the “robot” is a sibling, parent, or stuffed animal.

The DBE’s 2025 Foundation Phase rollout gives these ideas a clearer place in early learning. It does not mean every school will immediately have identical devices, teacher training, connectivity, or timetables. Families should expect uneven access and avoid treating access to equipment as a measure of a child’s ability.

For a wider view of how home learning can work when connectivity is unreliable, see this guide to offline-first education technology in Sub-Saharan Africa. Its central principle applies here: choose activities that still work when the internet does not.

Start with a realistic family plan

A sustainable plan is usually 30 to 60 minutes a week, plus ordinary conversations about technology. The aim is not to race ahead of school content. It is to help children become comfortable with trying, revising, and asking sensible questions.

Choose a regular time that does not compete with the busiest part of the week. For one family, that may be Saturday morning before errands. For another, it may be 30 minutes after supper on a night when the television stays off. A short routine that happens most weeks is more useful than a large project that creates stress and disappears after two sessions.

A four-part weekly routine

• 10 minutes: notice a system. Ask how a lift, traffic light, washing machine, shop queue, or phone alarm follows rules.
• 15 minutes: make an algorithm. Write or draw precise steps for a familiar task, such as packing a lunch or watering a plant.
• 15 to 25 minutes: build, test, and change. Try an unplugged challenge, a simple program, or a model.
• 5 minutes: reflect. Ask what failed, what changed, and what instruction was unclear.

Keep a small “design notebook.” Loose paper stapled together is enough. Each page can show the problem, first plan, test result, and improved plan. This makes progress visible even if a family shares one phone or has no home computer.

If writing is difficult for a younger child or a learner with handwriting, language, or attention needs, let them draw, dictate an answer, use picture cards, or explain their plan aloud while an adult records a few words. The thinking matters more than neat written work.

Choose the next task by age and confidence

Stage	Useful focus	A practical activity
Foundation Phase	Order, directions, patterns, cause and effect	Create arrow cards and guide a toy through a floor maze.
Intermediate Phase	Loops, conditions, variables, design constraints	Design a paper recycling sorter and write the rules it should follow.
Senior Phase and secondary	Algorithms, data, bias, privacy, testing claims	Compare two routes to solve the same task; explain which has fewer steps and why.

A confident older learner can help a younger sibling, but should not become the household teacher. Give each child a role: navigator, builder, tester, recorder, or error-finder. Rotate roles each session so the child who usually takes charge also has to listen and test someone else’s instructions.

Low-cost coding and robotics projects, including outage days

Power cuts and limited data do not need to stop learning. Unplugged work often makes abstract ideas easier to see. Prepare one envelope or box with paper, a marker, tape, string, recycled cardboard, bottle tops, scissors, and a few small objects. Supervise cutting and any project involving batteries, tools, or small parts.

1. The human robot maze

Time: 20 to 30 minutes. Cost: none.

Make a grid on the floor with tape, tiles, or sheets of paper. Put an object at one end and a “delivery point” at the other. The child may use only commands such as forward, turn left, turn right, pick up, and put down. An adult or sibling acts as the robot and follows the commands exactly, including mistakes.

The useful moment comes when the robot does the wrong thing. Say, “Let’s debug the instructions. Which step needs changing?” Do not silently correct the route. The child learns that an error is information, not a reason to give up.

For a child who finds a full maze overwhelming, begin with three squares and two commands. For an older learner, add a rule such as “repeat forward three times” or place an obstacle in the route after the first test.

2. A cardboard robot with a real job

Time: 45 to 60 minutes. Cost: recycled materials.

Ask the child to design a robot for a local or household problem: carrying books, reminding a family to save water, sorting recycling, or helping someone find a lost item. The model does not need motors. Its value is in the design brief and the decisions behind it.

• What problem does it solve?
• What information does it need to notice?
• What rule tells it what to do next?
• What could go wrong?
• Who might be left out or affected by its mistakes?

For example, a composite family project might be a cardboard “water-saving robot” that notices when a tap has been left running. A younger child can draw its buttons and warning light. An older child can write a flowchart: “If water is running and no one is using the sink, give a reminder.” Then ask what information the robot would need before it could make that decision reliably.

This leads naturally into AI literacy: computer systems act on information and instructions people provide. They do not automatically understand fairness, context, or a family’s values.

3. Sort, classify, and question the rules

Time: 15 to 25 minutes. Cost: none.

Gather safe household objects or picture cards. First, let a child sort them by one clear feature, such as colour or material. Then give an unclear instruction: “Sort the useful things.” Discuss why people could make different choices. A computer system also needs clear categories and examples, but unclear labels can produce weak or unfair results.

For older learners, add tricky examples. Is a reusable bottle “plastic,” “school item,” or “sports item”? The point is not to find one perfect answer. It is to see that categories reflect decisions made by people, and that a system can produce a poor result when its categories do not fit the real situation.

4. Device session: create, then explain

When a device and connection are available, choose a beginner-friendly block-based activity or a school-recommended platform. Set a 30-minute timer: 10 minutes to plan, 15 minutes to create, and 5 minutes to explain the logic aloud. Do not measure success by how colourful or complicated the screen result looks. Ask the learner to identify one instruction, one test, and one revision.

If a child becomes stuck, use a simple script: “Show me what you expected. Show me what happened. What is one small change we can test?” This keeps the task manageable and avoids taking over the keyboard.

Families looking for adjacent hands-on work can pair this with low-prep science experiment ideas. A bridge, ramp, or water-filter investigation becomes a robotics-style engineering task when children define a constraint, test a model, and record changes.

Teach AI literacy without treating a tool as an authority

AI literacy is broader than using a chatbot or image generator. A primary learner needs basic language for what a system can and cannot do. A secondary learner also needs habits of checking information, protecting privacy, recognising possible bias, and showing their own thinking.

A useful family rule is: AI can assist with ideas, but it is not a source to trust without checking. If a child uses an AI-enabled tool for schoolwork, ask them to identify what they used, check important claims against a reliable source, and rewrite the final answer in their own words. Follow the school’s current policy for assessed work.

Words a parent can use

• “What information do you think this system used to make that answer?”
• “How could we check whether that claim is correct?”
• “Would this work equally well for everyone? Why or why not?”
• “Show me your own plan before you ask a tool for suggestions.”
• “Do not paste your full name, address, school details, passwords, photographs of others, or private family information into a tool.”

A practical approach for older learners is to ask for a short planning outline before they use an AI tool. Afterward, ask them to mark which ideas came from the tool, which came from their own notes, and which claims still need checking. This is not about catching them out. It helps them distinguish assistance from understanding.

For a broader explanation of classroom and home uses, risks, and good questions to ask, read our family guide to AI in education. The goal is calm supervision rather than fear or unrestricted use.

Work around the digital divide with dignity

Digital access remains a national priority, but a family’s daily reality may include shared devices, mobile-only access, limited data, outages, or no reliable connection. A good home plan should not depend on uninterrupted streaming, paid subscriptions, or a parent who already knows programming.

• One shared phone: let children plan on paper first, then take turns entering or testing one section.
• Limited data: download teacher-approved material when connected; use screenshots and offline documents rather than repeated video viewing.
• No printer: draw arrow cards, grids, and flowcharts by hand.
• Several children: make a group project where each child owns one stage of the algorithm.
• Parent feels uncertain: ask questions instead of giving answers. “What should the robot do first?” is enough to begin.
• A child needs shorter tasks: offer one clear goal, such as moving a toy through three squares, then stop and return later.

Do not shame a child for not knowing a device function or compare home equipment with another family’s. Computational thinking is possible with a pencil, a route map, a recipe, or a box of bottle tops.

Use public DBE resources for revision and context

Begin with the DBE’s Coding and Robotics curriculum information to see official materials and updates. Resource pages can change, so check the current page and your child’s grade before relying on a document. Ask the school which activities or terminology it is using; this helps home practice support rather than duplicate classroom work.

For secondary learners, the DBE’s National Senior Certificate past examination papers are a useful public revision starting point for subjects that support digital and analytical skills. They are not a Coding and Robotics substitute, and younger children do not need them. Use them to practise reading instructions, planning time, and checking answers where relevant to the learner’s actual subjects.

For curriculum updates and policy developments, use primary DBE notices where possible rather than social-media summaries. Families who want occasional education-policy updates can also sign up for the education policy newsletter.

Common mistakes to avoid

• Buying equipment before building routines. A robot kit left unopened teaches less than four paper-and-cardboard challenges completed together.
• Equating coding with typing. Typing may be useful, but sequencing, logic, and revision are the deeper skills.
• Rescuing too quickly. Give a child time to test an idea. Ask what they observed before suggesting a fix.
• Using AI-generated output as finished schoolwork. This can hide gaps in understanding and may conflict with school rules.
• Ignoring privacy. Treat personal data, images, and school information as private unless a trusted adult has checked the setting and purpose.
• Making every activity digital. Outage-ready, hands-on work is not second-best; it is often the clearest way to learn the concepts.

FAQs

Does my Foundation Phase child need a laptop for Coding and Robotics?

No. Device access can help, but early learning can begin with directions, patterns, sorting, storytelling, and physical design challenges. Check with the school about its particular activities and resources.

What is the difference between coding, robotics, and AI literacy?

Coding is writing or arranging instructions for a computer or system. Robotics applies instructions to a physical device or model that performs a task. AI literacy is understanding how AI-enabled systems use information, where they can be useful, and why people must check outputs, protect privacy, and consider fairness.

How much time should a family spend each week?

Start with one 30-minute session for younger children or one 45- to 60-minute session for older learners. Consistency matters more than long sessions. Stop while the child still has enough energy to return next week.

Your next step this week

Set aside 30 minutes, make five arrow cards—forward, back, left, right, and stop—and run the human robot maze. Take one photograph of the final route or save the child’s drawing in their design notebook. Then visit the DBE Coding and Robotics page to identify one term or activity that connects with your child’s grade.