India's construction industry kills approximately 38 workers daily — and the helmet they're given hasn't meaningfully evolved since the 1980s. We redesigned the construction helmet as a smart, worker-centric safety system that senses, decides, and acts without burdening the worker.
Traditional construction helmets focus exclusively on impact protection — absorbing blows from falling objects. But the actual reasons workers die or get injured on Indian construction sites are far broader: heat exhaustion, proximity to hazard zones, improper wear, and zero real-time feedback. The helmet hasn't meaningfully changed since BIS introduced IS 2925 in 1984.
Worse — workers actively remove their helmets because they're uncomfortable. Heat-trapping shells, poor ventilation, one-size-fits-all fit, and neck strain from 8–10 hour shifts make abandonment a rational response, not negligence.
Safety equipment that is too uncomfortable to wear consistently is not safe equipment — it's a compliance checkbox. Workers treat helmets as fine-avoidance gear, not protection. The system fails not because of the helmet's impact rating, but because its design ignores the 8 hours between impacts.
We conducted primary research at active construction sites — observing workers in real conditions, interviewing them about their relationship with helmets, and mapping the gap between compliance expectations and actual behavior.
Watched how workers wear (and remove) helmets across shifts. Documented loosened straps, tilted positioning, complete removal during breaks, and the moment-to-moment decisions that lead to non-compliance.
Spoke directly with construction laborers about pain points — heat, weight, fit, sweat, neck strain. Discovered that workers see helmets as productivity inhibitors, not protection. The mental model is "avoid fines" not "stay safe."
Analyzed construction fatality data — falls from heights (33–59% of deaths), struck by objects (10–21%), electrocution (6.6–8.6%), structural collapses (~10%). Mapped primary causes to helmet design failures.
Deep-dive into BIS IS 2925, ISI certification requirements, material standards (HDPE, ABS, PVC), and the gap between regulatory mandates and on-ground enforcement reality.
Our research surfaced a constellation of insights that all point to the same conclusion: the helmet is not a product problem — it's a systems problem. Each insight maps to a specific design implication.
Workers remove helmets because existing designs prioritize compliance over long-duration comfort. Comfort is a prerequisite for safety adoption, not an add-on feature.
Thermal discomfort accumulates faster than perceived safety benefits. Helmet abandonment in heat is a rational response, not negligence. Reducing thermal load directly improves compliance.
A helmet worn incorrectly creates a false sense of safety for both workers and supervisors. Correct usage must be enabled passively, not enforced manually.
Workers operate in cognitively demanding environments. Adding complex "smart" interactions increases mental load. Intelligence should work silently in the background.
When safety tools resemble surveillance systems, trust collapses and adoption resistance increases. Worker dignity must be preserved.
Helmets, sensors, supervisors, and safety protocols function as disconnected entities. Safety must be approached as an ecosystem, not a standalone product.
Kavach operates on a six-stage smart loop that runs entirely on the helmet — no phone dependency, no cloud requirement, no internet needed. The system works with minimal training for workers with low digital literacy.
Helmet presence, proximity to hazard, sudden impact
Is helmet worn? In danger zone? Critical event?
Vibrate helmet, flash LED, trigger local alert
Worker corrects behavior, exits danger zone
Event logged locally, supervisor notified if critical
Works offline, defaults to physical alerts
The BIS-certified helmet shell remains unchanged. We're not redesigning the helmet — we're adding a smart layer on top of it. The system must work with existing IS 2925 compliant shells, require no structural modification, and add negligible weight. Fixed assumptions: no internet, low digital literacy, minimal training, zero phone dependency.
Detects whether the helmet is properly worn. If removed or tilted, haptic vibration + LED alert triggers. No manual check needed — passive compliance enforcement.
Multi-modal alerts (vibration + visual LED) for hazard proximity, improper wear, and zone breaches. Designed to work in noisy environments where audio alerts fail.
Detects when a worker enters a designated hazard zone. Triggers escalating alerts based on proximity — gentle vibration at perimeter, intense at threshold.
One-press SOS for emergencies. Temperature-responsive ventilation adjusts airflow based on thermal conditions. Modular accessory racks for site-specific add-ons.
Helmet-to-helmet mesh network operates without WiFi or cellular. Supervisor-level notifications propagate through the mesh. Events logged locally and synced when connectivity returns.
We built a working prototype with sensors mounted on a BIS-certified helmet shell. The electronics integrate without structural modification — the smart layer clips onto the existing suspension system and adds minimal weight.
Prototype images coming soon.
Led primary research at construction sites — observations, worker interviews, and behavioral mapping. Compiled and analyzed accident data from NCRB and industry reports. Synthesized 11 key insights that drove every design decision.
Designed the sensor architecture, selected components (ESP32, IMU, temp/proximity sensors, haptic motors), built the working prototype, and programmed the Sense → Decide → Act → Fail-safe loop in Arduino IDE.
Led concept exploration and selection — evaluated multiple form factors against feasibility, user comfort, and BIS compliance. Defined the modular add-on system, ventilation approach, and SOS interaction.
Spearheaded a team of 5, set direction, managed scope within the 2-week sprint, and delivered the final presentation to faculty — connecting every feature back to a specific research insight.
Kavach was presented to faculty and well received — recognized for its depth of field research, the insight-to-design traceability, and the working prototype that demonstrated the smart loop in real hardware. The project proved that construction safety isn't a helmet problem — it's an ecosystem problem requiring worker-centric systems thinking.
The project continues with three priorities: refining the form factor for construction compliance, adding more passive-dependent technology to avoid connectivity issues, and aligning the system closer to workers' existing mental models — making it invisible enough that they forget it's there.