Nagarika turns Ward 23 of Hosur into a live map where the ward's real garbage and sanitation problems become the game's objectives. For this pilot it runs as the game layer of CMCA's HCAP programme, played by school students and their parents.
Clean-up drives work, but interest fades after each one and impact is hard to measure. The pilot tests whether a game keeps families taking action between drives.
What the pilot must prove
≥25%
of registered student players are still active in the last live week (19–25 Nov).
Active means at least one verified action that week. This is the pilot's pass/fail test. Diwali week (5–11 Nov) is recorded but not counted against the pilot, since schools are on holiday.
Scope
- Where
- Ward 23, Hosur.
- Who plays
- Students from HCAP schools in Ward 23 or nearest to it, with their parents or guardians. Target about 100 students.
- Launched by
- CMCA and A2Z together, as part of the HCAP programme.
- Delivery
- Two options, both shared through CMCA and school WhatsApp groups: a web link that opens in the phone's browser and can be saved to the home screen, and an Android app file (APK). No Play Store for the pilot.
- Escalation
- Simulated in the game. No mails go to HMC or the ward office during the pilot.
- Duration
- 2 weeks pre-pilot (15–28 Oct) + 4 weeks live (29 Oct – 25 Nov), then a review week.
What students and parents do in the app
A mix of four activity types, rotated across the four live weeks.
Segregation check-ins
Log wet / dry / plastic sorting at home with a parent. Builds the daily habit.
Report spots, with an adult
Photograph garbage spots on the Ward 23 map: Blackspots (recurring dumps) and Flashpoints (one-off litter). Always with a parent or guardian.
Awareness journeys
Mini-games on what happens to garbage after it leaves the bin. Finishing one unlocks in-game tools, like Voice (file a complaint) and Beacon (mark a spot).
Family ripple tasks
Weekly challenges done as a family, such as a clean-street challenge in Diwali week.
How it works in the pilot
| Component | Pilot setup | Status |
|---|---|---|
| Accounts | At login, the player picks student or guardian. Everyone logs in with a username and phone number, confirmed by a one-time SMS code. Students use their own phone number if they have one, otherwise a parent's. Whether a parent approves each student profile is still to be decided. | In build |
| Checking reports | Automatic. The phone's location must be inside Ward 23 and near the spot, and an AI model checks the photo: is it garbage, is it a repeat, is it safe to show? Reports it flags are held for review by a person (reviewer to be decided). | In build |
| Map | Map served from our own server. Ola Maps is built in as a backup but switched off. Google Maps dropped. | Decided 19 Sep |
| Login SMS | Login codes sent through an Indian SMS provider (MSG91 or Gupshup). Needs our sender name registered with TRAI first (DLT registration). | Decided 11 Aug |
| Languages | Tamil and English in the pilot, including AI-written spot descriptions. | Tamil in progress |
| Rewards | Leaderboards by ward, by area and by school; recognition at school; small prizes for top players. | Prizes TBC |
Pilot objectives
- Return: ≥25% of registered students active in the last live week. This is the pass/fail test.
- Reach: about 100 students registered.
- First use: most new players do their first check-in or report within 48 hours of signing up.
- Real reports: every report is a verified photo pinned to a real place on the Ward 23 map.
- Safety: no inappropriate photo is shown publicly. Whether every student profile needs a parent's approval is to be decided.
- Check accuracy: compare the automatic checks with a person reviewing a sample, to see how often they get it right.
- App format: see whether families use the web link or the APK more.
- Map cost: count how much map data each visit loads. This tells us whether paying for Ola Maps later would be affordable.
Timeline
W1 · Pre15–21 Oct
Schools, parent consent, SMS registration, Tamil
W2 · Pre22–28 Oct
App finalised, trial with 10–15 families
W3 · Live 129 Oct–4 Nov
Launch, onboard, check-ins
W4 · Live 25–11 Nov
Diwali family challenge
W5 · Live 312–18 Nov
Garbage spot hunt · Children's Day
W6 · Live 419–25 Nov
Fix-it round · final, counted week
Review26 Nov–2 Dec
Data, report, decision
Pre-pilotLive pilotReview
The pre-pilot is two weeks, and registering the SMS sender name (DLT) alone can take one to two weeks. So the pilot agreement and the SMS registration both start before 15 Oct.
Milestones
| # | Milestone | Owner | Target | Status |
|---|---|---|---|---|
| M1 | Pilot agreement signed with CMCA (roles, schools, contacts) | Madhav | 10 Oct | In discussion |
| M2 | SMS sender name registration (DLT) filed | Ranga To confirm | 10 Oct | Pending |
| M3 | Participating schools confirmed, one teacher contact per school | Madhav + CMCA | 17 Oct | Pending |
| M4 | Parent consent screens and privacy check done | Madhav + Ranga | 21 Oct | Pending |
| M5 | App locked for launch: student/parent login, automatic checks, Tamil, web link + APK | Ranga + Srivathsa | 24 Oct | Pending |
| M6 | Trial run with 10–15 families, fixes made | All | 28 Oct | Pending |
| M7 | Pilot launch in Ward 23 | Madhav + CMCA | 29 Oct | Planned |
| M8 | Mid-pilot check: who is still playing, wrong checks, support issues | Madhav | 12 Nov | Planned |
| M9 | Pilot ends, data pulled | Ranga | 25 Nov | Planned |
| M10 | Final review and decision with CMCA | Madhav | 2 Dec | Planned |
Week-by-week deliverables
| Week | Product & deployment (Madhav) | Tech (Ranga, Srivathsa) | CMCA / schools |
|---|---|---|---|
| Before 15 Oct | Share pilot design with CMCA. Shortlist schools together. | File SMS sender registration (DLT). Plan the student/parent login and automatic checks. | Nominate pilot lead. Share HCAP school list for Ward 23. |
| W1 · 15–21 Oct | Parent consent text and privacy checklist. List of content to translate into Tamil. Script for teachers to introduce the app. | Build student/parent login, location check and AI photo check. Log how much map data each visit uses. | Introductions to schools. One teacher contact per school. |
| W2 · 22–28 Oct | Trial with 10–15 families. WhatsApp starter pack: link, APK and a how-to in Tamil. Leaderboard and prize rules. | Lock the app on 24 Oct. Fix trial bugs. APK and web link ready. | Schedule onboarding sessions in schools. Share links to parent groups. |
| W3 · 29 Oct–4 Nov | Launch. Track sign-ups and first actions daily. Answer questions in the support group. | Watch for errors and wrong checks. Fix problems the same day. | Onboarding sessions in class. Segregation check-ins start. |
| W4 · 5–11 Nov | Diwali clean-street family challenge. Expect lower activity; send WhatsApp reminders. | Turn on festival-litter spots on the map. | Push the challenge to parent groups before the holiday. |
| W5 · 12–18 Nov | Mid-pilot check on 12 Nov. Garbage spot hunt with parents. Children's Day (14 Nov) recognition. | Release the second awareness game. Update leaderboards. | Shout-outs at school for top players. |
| W6 · 19–25 Nov | Final week, the one the 25% test is measured on. The top spots play out the full story in the game: reminder, escalation, cleaned up. No real mails are sent. Final leaderboard. | Pull the data and build the results summary. | Prize and recognition event. |
| Review · 26 Nov–2 Dec | Pilot report: did we reach 25%, what we learned, map cost. Decision meeting. | What broke, what to change, cost for more users. | Feedback from teachers and field team. |
Team
| Person | Organisation | Role in the pilot |
|---|---|---|
| Madhav | A2Z | Product and project manager. Deployment, CMCA and school coordination, metrics, go/no-go report. |
| Ranga | A2Z | Tech lead. Builds and runs the app, report checks, data. |
| Srivathsa | A2Z | Tech. Build, fixes, support during the live weeks. |
| CMCA pilot lead To be nominated | CMCA | Leads the pilot on CMCA's side. Signs off the design and the final decision. |
| Hosur field team To be nominated | CMCA | Introduces us to schools, runs app introduction sessions, first point of help for families. |
| Teacher per school To be nominated | Schools | Introduces the app in class, shares links with parents, recognition. |
Phase
Planning
Pilot design shared with CMCA for review.
Next milestone
M1 · Pilot agreement signed
Target 10 Oct
Days to pre-pilot
14
Pre-pilot starts 15 Oct
What we need from CMCA
| Item | Needed by | Status |
|---|---|---|
| Review and sign off the pilot design | 10 Oct | Pending |
| Nominate a pilot lead and a Hosur field team contact | 10 Oct | Pending |
| Confirm participating HCAP schools in or nearest to Ward 23, with introductions | 17 Oct | Pending |
| One teacher contact per school | 17 Oct | Pending |
| Class time to introduce the app (week of 29 Oct) | 24 Oct | Pending |
| Share the app with parents through school and HCAP WhatsApp groups | 24 Oct | Pending |
| Recognition moments at school (assembly, Children's Day, closing event) | 7 Nov | Pending |
Communication
| What | Who | When |
|---|---|---|
| Pilot review: progress, numbers, blockers | Madhav + CMCA pilot lead | Weekly |
| Field issues and app problems | A2Z team + CMCA field team | WhatsApp group, as they come up |
| Single point of contact for CMCA | Madhav | Throughout |
Status log
| Date | Update |
|---|---|
| 17 Aug | Nagarika demo at Friends of CMCA. |
| 19 Sep | Map and hosting approach finalised. |
| 1 Oct | Pilot design shared with CMCA. |
| W1 | Pending |
| W2 | Pending |
Risks
| Risk | Impact | Mitigation |
|---|---|---|
| Short pre-pilot window: SMS sender registration, Tamil content and consent screens all due in two weeks | High | SMS registration filed before 15 Oct. Trial run on 27–28 Oct catches gaps before launch. |
| Children's data and parent consent | High | Students may log in with their own or a parent's phone number. Whether a parent approves each student profile is to be decided. Privacy check before launch. |
| An unsuitable photo gets through the automatic check | Medium | Photos stay hidden until they pass. Flagged reports are held for review. |
| Diwali holidays lower activity in week 4 | Medium | Diwali family challenge. Week 4 is not counted in the 25% result. |
| Families hesitate to install an app file (APK) | Medium | Web link offered first. APK shared only through school and CMCA groups. |
| Siblings share one phone | Medium | Activity is counted per student, not per phone. |
| Phone location is off or inaccurate | Medium | Reports accepted within a small distance of the spot. Failed checks are held for review, not rejected. |
| Fewer sign-ups than expected | Medium | App introduced in class by teachers. School and area leaderboards. Sign-ups tracked daily in week 3. |
Assumptions
- Nagarika reaches families through HCAP schools. Teachers introduce it in class and share it in parent groups. If it goes out another way, sign-ups, the asks of CMCA and the timeline all change.
- CMCA nominates a pilot lead, a field team contact, and one teacher per school
- Participating schools give class time to introduce the app
- Most families have at least one Android phone
- Parents are reachable through school WhatsApp groups
Decisions
- Pilot covers Ward 23 only, as part of HCAP
- Players are HCAP students and their parents
- Complaints and escalation happen inside the game only; no mails go to HMC during the pilot
- App shared as a web link and an APK; not on the Play Store
- Reports checked automatically using location and AI
- Madhav is the single point of contact for CMCA
- Next step decided together at the final review
Pilot agreement
- Between
- A2Z Antifragility Foundation × CMCA
- Pilot
- 6 weeks · 15 Oct – 25 Nov 2026 · Ward 23, Hosur
- Target sign-off
- By 10 Oct 2026
- Roles
- CMCA introduces Nagarika to schools and parents through HCAP. A2Z builds, runs and supports the app. This assumes the game reaches families through HCAP schools.
- Who pays
- To be agreed in the pilot agreement.
- Covers
- Roles and contacts, pilot costs, data ownership, and how photos are stored and for how long.
Why this matters for HCAP
- Regular, measurable activity between clean-up drives, not just during them.
- Every action leaves a record: a photo, a place on the map, a date.
- A working model for the playbook's Year 2–3 goals: better local reporting and a citizen complaint system.
Pilot costs
Two versions: the 6-week pilot on this portal (15 Oct – 25 Nov), and the original 12-week estimate.
| Item | What it covers | 6 weeks | 12 weeks |
|---|---|---|---|
| One-time setup | Accounts, SMS sender registration, domain and setup | ₹9,860 | ₹9,860 |
| Running cost | Hosting, database, backups, SMS, AI-written descriptions | ≈ ₹9,500 | ≈ ₹19,000 |
| Base total | Before items marked TBC | ≈ ₹19–20k | ₹28,905 |
| AI photo checks | Checking each photo before it appears on the map | TBC | TBC |
| Prizes | Small prizes for top players and schools | TBC | TBC |
| Tamil translation | App text and the WhatsApp starter pack | TBC | TBC |
| Privacy check | Consent text and data handling | TBC | TBC |
| Other costs | Such as a reviewer's time, onboarding materials, the prize event and field travel | TBD | TBD |
Success criteria · Go / No-Go
| Criterion | Target |
|---|---|
| Students still active in the last live week (19–25 Nov) | ≥ 25% |
| Student profiles approved by a parent or guardian | 100% To confirm |
| Unsuitable content shown publicly | 0 |
| Students registered | ≈ 100 |
| New players who act within 48 hours of signing up | ≥ 60% |
| Automatic checks that match a person's review | ≥ 90% |
| Web link vs APK use | Measured |
| Map data loaded per visit (sets the cost of a wider launch) | Measured |
After the pilot
At the final review on 2 Dec, A2Z and CMCA decide the next step together, based on the 25% result and what the pilot shows. Options include expanding to more HCAP wards, a Hosur-wide launch with CSR partners, a second phase in Ward 23 with real complaints to HMC, or a revised re-run.