Direction 01
A casting-defect spotter for foundries: photograph parts, learn the shop’s specific defect patterns, catch drift early
Sustainable Development Goal 09 / 17
Industry, Innovation and Infrastructure
Build resilient infrastructure, promote inclusive and sustainable industrialisation and foster innovation. For builders: tools that modernise MSMEs, maintain public infrastructure and democratise industrial intelligence.
Why this goal matters
The technology gap inside economies is wider than the gap between them. A large factory has sensors, dashboards and planners; the workshop across the road runs on memory and paper. Public infrastructure — roads, bridges, buses — decays quietly until failure. Software that brings industrial-grade intelligence to small operators and public assets multiplies productivity where it matters most.
The Indian context
MSMEs contribute a huge share of India’s manufacturing output and employment yet remain the least digitised layer of the economy. National infrastructure investment is at record levels, which makes maintenance intelligence — knowing what is degrading where — an increasingly valuable public capability.
Tamil Nadu, specifically
Coimbatore is one of India’s great MSME cities: pumps, motors, foundries, machining shops — often family-run, deeply skilled and minimally digitised. Chennai anchors automotive and electronics manufacturing corridors. Between them sits a living laboratory for shop-floor AI that respects how small industry actually works.
People and communities affected
- MSME owners and shop-floor workers
- Commuters dependent on ageing public transport
- Residents affected by infrastructure failures
- Industrial-cluster job seekers
- Local bodies maintaining civic assets
Key challenge areas
- Shop-floor digitisation without disruption
- Predictive quality and maintenance for small factories
- Public-asset condition monitoring
- Supply-chain visibility for clusters
- Skilled-trade knowledge preservation
The AI Lens
Where AI belongs here — and where it does not.
Where AI can meaningfully help
- Detecting quality defects from phone-camera images at a fraction of machine-vision-system cost
- Predicting machine failures from vibration and sound patterns on affordable sensors
- Reading handwritten job cards and registers into structured production data
- Prioritising road or asset repairs from citizen photo reports and inspection history
Where AI may be inappropriate
- Fully autonomous control of dangerous machinery without certified safety engineering
- Worker-replacement framing in communities where the promise should be augmentation
- Infrastructure-risk scoring published without engineering validation — panic has costs
Build Directions
Example problem directions
Starting points, not prescriptions — the strongest submissions narrow one of these into a specific, evidenced local problem.
Direction 02
A machine-health listener that learns each machine’s normal sound and flags deviation
Direction 03
A pothole-and-asset reporter that deduplicates citizen photos and builds repair-priority maps
Direction 04
A digital job-card system that speaks the shop floor’s language, literally
Potential users
- MSME owners and cluster associations
- Maintenance technicians
- Municipal engineering departments
- Industrial-training institutes
Possible datasets
- Open industrial defect and machinery-sound research datasets for pre-training
- Your partner workshop’s own images and readings, with consent
- Municipal complaint and works data where published
- MSME ministry cluster statistics (public, aggregate)
Data-access limitations
- Every shop’s defects and machines differ — plan for per-site fine-tuning
- Industrial data is commercially sensitive; agreements matter
- Public asset data is fragmented across departments
Privacy, bias and safety risks
- A missed defect that ships erodes trust instantly — set thresholds honestly
- Sensor projects die without maintenance plans
- Digitisation that adds work for the shop floor gets abandoned
Responsible-AI considerations
- Fit the tool to existing workflows; do not demand new ones
- Keep the shop’s data the shop’s property
- State detection limits explicitly; never imply certified inspection
Suggested impact metrics
- Defect-catch rate versus current practice in a pilot line
- Downtime avoided through early warnings, measured against history
- Paper-to-digital time savings per job card
Related SDGs
Related challenge areas
Participant resources
Build for Industry, Innovation and Infrastructure.
Anchor your team in this goal, define one real problem, and start the three-lock journey.
Register Your Team