01 / The moment we miss

A busy street. An unseen emergency.

How might we spot an injured animal quickly?

Everyone is moving. Help has not been alerted.

Urban animal welfare · Make noticing the first step.

Animated concept illustration. No real animals or reported incidents depicted.

02 / Problem identification

The barrier is bigger than awareness.

Sick, distressed and hit-and-run animals can be seen, yet still go unreported.

60.4%keep driving. Respondents say they ignore the animal because they fear order delays. Current behaviour · n = 91
  • Delivery timers / app penalties54.4%49 of 90
  • No emergency reporting sites44.4%40 of 90
  • Reporting takes too much time26.7%24 of 90

Source: supplied delivery-partner survey screenshots. Reasons question n = 90; multiple selections. No citywide injury-per-day estimate was supplied. People should not have to take on medical care to raise an alert.

03 / The people already on the street

91 responses. A clear design signal.

Delivery partners from Blinkit, Swiggy and Zepto, as described by the survey team.

71.5%encounter injured animals at least a few times a week35.2% daily + 36.3% weekly · n = 91
63.3%choose snack tokens as their strongest motivationtea, coffee or samosa · n = 90
71.1%prefer a quick photo with GPS64 of 90 · multiple selections
62.2%want a reporting time under 30 seconds56 of 90 · 23.3% (21) between orders

Not a representative city census; validates a concept, not population-wide prevalence. Question totals vary from 89 to 91. Source: four supplied Google Forms screenshots. 71.5% is the sum of displayed rounded values. Company names identify the proposed network, not confirmed partnerships.

04 / A new response

One safe stop. One photo. A chance to help.

Riders are spotters, not care providers.

  1. 01 Stop safely
  2. 02 Photo + GPS
  3. 03 Alert care team
  4. 04 Help arrives, status tracked

Illustrated proposed workflow. Alerts and care are simulated, not a live emergency service. Never use a phone while riding.

05 / Reporting interface

Small action. No medical responsibility.

Target: a sub-30-second report, once safely stopped.

  • 1-tap photo with auto-detected GPS tag
  • Optional voice note or map pin, only when safely stopped
  • Location and visual evidence are enough. No handling, treatment or waiting with the animal.

64% fear late-delivery penalties or complaints (57 of 89, multiple selections). A no-penalty policy would need delivery-platform agreement. It is a proposal, not a guarantee.

Ready

Concept interface. The demo uses no camera, sends no real alert and collects no location.

06 / Three-tier platform

A report becomes a trackable action.

One reporting network. Three connected sections.

  1. Reports Report history, timestamps and live status: submitted, verified, assigned, closed.
  2. Rewards Verified reports earn points, redeemed for tea, coffee or samosa at participating local vendors.
  3. Shift incentives An LP-based engine proposes shift-level bonus payouts within an agreed daily budget.

Verification, duplicate checks and a responder network are needed before a report becomes care. Proposed architecture; delivery-platform and vendor integrations are not established.

07 / The allocation question

Put a daily budget where it matters.

Linear programming chooses bonus credits per verified report for each shift.

B max
Daily funding or points limit
p
Expected active-driver participation, e.g. 10%, 20%, 40–50%
Et
Expected verified reports per shift, from participation and exposure
rt
Explicit risk weights; calibrate before use

Four windows: morning 6 am–12 pm, afternoon 12–5 pm, evening peak 5–10 pm, late night 10 pm–4 am. 4–6 am is outside the supplied definition. Survey shift answers overlap, so they are not a driver-density distribution. Shift risk and volumes need operational data.

08 / LP formulation

A linear model, with visible guardrails.

Hold expected report counts fixed per scenario. Choose ct, the bonus per verified report.

Maximise risk-weighted credit scoreZ = Σ rt · ct · Et
BudgetΣ ct · Et ≤ Bmax
Motivation floorct ≥ ₹15–₹20
Practical capct ≤ ut

The objective weights spend by shift risk; it does not prove higher payouts cause more reports or rescues. Without caps, all spare budget could land in one shift, so caps are an explicit modelling addition. The graph is an illustrative two-shift slice.

09 / Explore the LP

What does the budget actually buy?

An illustrative solved scenario, not field results. Move the budget and watch the allocation.

₹1,000₹3,600

80 expected reports · ₹20 minimum · ₹40 cap · risk weights 1, 1.2, 1.8, 2.2 · counts 15, 15, 30, 20. All inputs are synthetic.

Bounded continuous LP: extra budget goes to the highest risk weights first, subject to caps. ₹20 is a proposed floor, not a proven optimal payout. No predicted uplift is claimed.

10 / From concept to pilot

Turn everyday routes into a safety net.

Faster reporting, a protected rider, and an accountable care response.

  1. 01 Start local: one service area, verified care partners and snack-voucher vendors.
  2. 02 Measure the chain: reporting time, valid reports, duplicates, response time, budget use.
  3. 03 Scale only after proof. App integrations with Swiggy, Blinkit, Zepto or Zomato are a proposed route, not a current partnership.

Spot. Report. Let care reach them.

Intended impact, not measured results. Download editable source ↗