India’s New Safety Mandates: Driving OEMs to Rethink ADAS Integration

India's automotive sector stands at a decisive regulatory inflection point. The Ministry of Road Transport and Highways has mandated a new suite of active safety technologies governed by specific Automotive Industry Standards: AIS 184 for Driver Drowsiness and Attention Warning, AIS 186 and AIS 187 for Blind Spot Detection and Moving-Off Vulnerable Road User Detection, AIS 188 for Lane Departure Warning, and AIS 162 for Automatic Emergency Braking.

With approximately 18 months to the compliance window, OEMs face serious engineering and architectural decisions. AIS 184, 186, 187, and 188 are alerting standards — they warn the driver and leave the decision with the human. AIS 162 demands active safety: autonomous vehicle actuation that overrides human control. That distinction changes everything.

Alert Quality: The Foundation of Driver Trust

Before addressing autonomous braking, the alerting standards deserve scrutiny — because their effectiveness directly conditions driver trust across the entire safety system. Across AIS 184, 186, 187, and 188, the shared risk is alert fatigue. A drowsiness warning that fires prematurely, a lane departure alert triggered by a deliberate merge, or a blind spot notification from a stationary roadside structure — each false positive erodes confidence and desensitises response to alerts that genuinely matter.

The right alert at the right time is a safety requirement, not a convenience specification. Achieving it demands contextual intelligence: assessing whether a vehicle in the blind spot poses an actual risk given the host vehicle's path, speed, and traffic state — a level of reasoning isolated sensors cannot deliver.

Fusion as Architecture, Not Feature

Treating each AIS mandate as a discrete system with its own sensor pipeline is the most consequential deployment mistake anyone can make. Fragmented modules cannot share context, resolve conflicting inputs, or prioritise the most critical threat when multiple conditions overlap.

Effective ADAS — and viable AEB specifically — requires a unified intelligence architecture: surround-view perception fused into a shared environmental model, processed on a single compute platform with latency suited to safety-critical decisions. Driver monitoring must feed every system, not operate as a standalone AIS 184 module, because driver attention state should modulate when every alert fires and at what threshold AEB intervenes. This is a systems engineering philosophy — the only approach that produces coherent, trustworthy behaviour on a real road.

Why AIS 162 Is an Entirely Different Problem

India's expanding expressway and national highway network carries high-speed, high-density traffic — compressing braking time windows and raising the stakes of both missed detections and false activations. The host vehicle may be a 20-metre rigid truck, an 80-metre wind turbine blade transporter, or a tanker carrying 25,000 litres of liquid cargo. Load weight, centre of gravity, fluid surge, and trailer articulation all interact with braking force in ways a uniform AEB calibration cannot accommodate — an intervention that prevents a collision in one loading state can induce a rollover in another.

Monsoon-reduced tyre adhesion and seasonal sensor degradation further complicate the decision envelope. The AI underpinning AIS 162 must be trained on this full complexity through real-world field data. Models built on foreign test tracks carry systematic blind spots in India's traffic mix and driving behaviour. Local data maturity is a baseline requirement, not a premium differentiator.

The Cost Trap: When Compliance Replaces Capability

Cost sensitivity is structurally embedded in Indian OEM procurement. When a mandate arrives, the instinct is to find the lowest-cost component that satisfies certification, or to pressure suppliers into downgrading specifications. For active safety, this is dangerous. A system that passes homologation under controlled conditions but performs inconsistently in the field actively undermines safety.

Drivers who experience chronic nuisance alerts will distrust the technology; an AEB system that activates incorrectly at highway speeds creates hazards. A homologation certificate and a system that reduces fatalities are not the same outcome — and deploying technology to tick a compliance box will never achieve the latter.

India's ADAS mandates define the minimum. Whether AIS 184, 186, 187, 188, and 162 translate into fewer collisions depends on implementation quality — contextually intelligent alerting systems that earn driver trust, and locally validated AI with full vehicle dynamics awareness for AEB. Compliance is the floor. Safety is the goal. OEMs that internalise that distinction will build systems that matter on the road.