Class 7 · Unit 3: AI at work · Lesson 3.4 · 40 min
AI in transport and communication
The highway's closed near Ramban. A text says you won ₹50,000 — and the phone has already binned it as spam. Two jobs AI does millions of times a day.
Today you will: Build a route rule with IF · Label messages and find the spam clues · Tell a written rule from a learned pattern
Story
Bilal's family is driving from Srinagar to Jammu for a wedding. At breakfast, his father checks his phone.
Bilal: “The highway is closed near Ramban. Landslide. The traffic police advisory says don't travel today.”
Bilal: “How does the map app know how long the drive will take on other days?”
Hiba, who is visiting, has her own phone problem. A message says: "CONGRATULATIONS! You won ₹50,000. Click this link now!" The phone has already put it in a Spam folder.
Route-planning and spam-catching are two jobs AI does millions of times a day.
Watch
AI on the road and in your inbox — — scan code 7.3.4 in the printed book.
Warm up your fingers
Skill: numbers and symbols inside sentences, Term 2 (7 min)
- On typing.com, practise a numbers row review lesson.
- Then type exactly:
NH-44: closed at Ramban. Travel time 8 hrs 30 min. - Watch the hyphen, the colon and the full stops.
On the laptop
You need: LibreOffice Calc · route-rule.ods (10 made-up trips) · messages.ods (12 made-up texts)
Mission 1: The route rule (pairs, 10 min)
- ☐1
Open
route-rule.ods: Day · Snow last night? (Y/N) · Advisory (Open/Closed) · Leave time · Decision. - ☐2
In E2, build a rule with IF:
=IF(C2="Closed";"Don't travel";IF(B2="Y";"Travel with care";"Travel")). Fill it down. - ☐3
Talk: this is a rule someone wrote. A map app learns travel times from data. Which one would know about today's landslide?
Mission 2: Spam or not spam? (pairs, 8 min)
- ☐4
Open
messages.ods. Predict: how many of the 12 are spam? Note it. - ☐5
Label each
SpamorNot spam, and in Clue type the feature that decided it — prize, link, urgent, unknown number. - ☐6
Which clue appears most in spam? That's the pattern a spam filter learns. Check the teacher's key: did you call a real message spam?
No laptop today?
With the 10 trips and 12 messages printed, pairs write each decision and label on paper; the teacher tallies the most common spam clues on the board.
Now you know
- Map apps predict travel time from past journeys. Predicting a number is regression.
- Spam filters classify messages using features — certain words, links, unknown senders.
- Wrong both ways: a real message blocked, or a scam let through.
- Official advisories come first. For highway closures and avalanche warnings, follow J&K Traffic Police and the district administration.
- Never click links you didn't expect — even if the filter let them through.
Debate it
A spam filter learned mostly from English messages. Scam messages in Urdu or Kashmiri keep getting through to people's phones.
Who is hurt most by this? What would the filter need to learn from?
Check yourself — practice, not a test
1. A map app predicting "8 hours 30 minutes" to Jammu is an example of…
- ○ Classification
- ○ Regression
- ○ Clustering
- ○ Sorting
2. A spam filter sorting messages into "spam" and "not spam" is an example of…
- ○ Classification
- ○ Regression
- ○ Clustering
- ○ Drawing a chart
3. Which features might make a spam filter suspicious? (choose all)
- ○ The word "WINNER"
- ○ A strange link
- ○ A message from your teacher's saved number
- ○ "Reply URGENTLY"
4. If a map app shows the highway as open, you can ignore the traffic police advisory that says it is closed.
- ○ True
- ○ False
Remember
**AI predicts the road and filters the inbox, and for safety, trust official warnings over any app.**