B8 · Computational Thinking and the AI Project Cycle

Problem Scoping and the 4Ws Canvas

Problem scoping is the first stage of the AI project cycle: understanding a problem deeply and describing it clearly before trying to solve it. The 4Ws canvas is a tool for this. It asks Who has the problem, What the problem is, Where it happens, and Why solving it matters, and ends with a one-sentence problem statement.

10 min Beginner Jump to the activity

What you need: Nothing. This page starts from the beginning, with no programming and no formulas.

Definition

"Students are unhappy with the canteen." Is that a problem you can solve? Not yet. Unhappy about what? Which students? When? Why does it matter?

Problem scoping is the first stage of the AI project cycle: understanding a problem deeply and describing it clearly before trying to solve it. The 4Ws canvas is a tool for this. It asks Who has the problem, What the problem is, Where it happens, and Why solving it matters, and ends with a one-sentence problem statement.

In simple words

A doctor does not hand you medicine the moment you walk in. They first ask questions: where does it hurt, since when, what makes it worse. Only when they understand the problem do they decide the treatment. Problem scoping is asking those questions before building anything.

Why it matters

A clever solution to the wrong problem helps nobody. Time spent understanding the problem saves far more time later, and keeps the project focused on the people who actually need help.

Key terms

Problem scoping
Understanding and clearly defining a problem before solving it.
4Ws canvas
A template with four boxes: Who, What, Where and Why.
Stakeholder
Anyone who has the problem, is affected by it or by its solution.
Evidence
Facts showing the problem is real, such as survey results, records or observations.
Root cause
The underlying reason a problem happens, not just its visible symptom.
Problem statement
One clear sentence describing who has the problem, what it is, where it happens and what an ideal solution would do.

Explained in detail

The four boxes

BoxQuestions to askCanteen example
WhoWho has the problem? Who else is affected? What do we know about them?Students who arrive in the second lunch break; canteen staff
WhatWhat exactly is the problem? What evidence shows it is real?Popular items sell out before the second break; a survey of 120 students found 70 per cent affected
WhereWhere and when does it happen? In what situation?The school canteen, mostly on days with sports practice
WhyWhy does solving it matter? Who benefits, and how?Students skip lunch and feel tired; the canteen wastes food it cooks too much of

Writing the problem statement

The four boxes are then joined into one sentence, often following a template: "Our [who] have a problem that [what] when [where]. An ideal solution would [why]." For the canteen: "Our students in the second lunch break have a problem that popular items run out before they arrive, on sports-practice days. An ideal solution would help the canteen predict demand so that everyone can buy lunch and less food is wasted." A clear statement like this guides every later stage: what data to collect, what to predict and how to judge success.

Symptoms and root causes

"Students are unhappy" is a symptom. Asking "why?" several times usually leads to the root cause. Why are they unhappy? Food runs out. Why? The canteen cooks the same amount every day. Why? It has no way of knowing how many students will come. Now there is something a data-based solution could address: predicting demand. This technique, asking "why" about five times, is widely used in industry and is sometimes called the five whys.

Choosing problems worth solving

  • Is the problem real, with evidence, or just an assumption?
  • Does it affect enough people, or affect a few people seriously?
  • Can data help, and can that data be collected fairly?
  • Have the people affected been asked what they need?

Many AI projects for social good start from the United Nations Sustainable Development Goals, such as zero hunger, good health, quality education and clean water, which provide a list of important problems that affect millions of people.

How it works

  1. Talk to the people who have the problem.
  2. Fill in Who, What, Where and Why, with evidence for each.
  3. Ask "why?" several times to reach the root cause.
  4. Write a one-sentence problem statement.
  5. Check it with the stakeholders: is this really their problem?

Try it yourself

From "unhappy" to a clear problem

A student team is asked to "fix the canteen".

Which W does it answer?

Sort each note from a project on village water supply.

  • The hand pump breaks down about twice a month and takes a week to repair

  • Three hamlets on the edge of the village, in summer

  • Girls miss school while walking to a distant well

  • Women and girls who fetch water each morning

Common mistakes

  • Starting with a solution ("let's make an app") instead of a problem.
  • Describing a symptom without finding the root cause.
  • Writing the problem without asking the people who have it.

Check yourself

Question 1

1. What are the 4Ws?

Question 2

2. Which is a good problem statement?

Question 3

3. Why ask "why?" several times?

Summary

  • Problem scoping defines the problem before solving it.
  • The 4Ws canvas asks Who, What, Where and Why, with evidence.
  • Find the root cause and write a one-sentence problem statement.

Finished reading?

Sources: CBSE Artificial Intelligence curriculum resources · United Nations Sustainable Development Goals