Sources: Apple plans a “crash detection” feature next year for iPhones and Watches, which measures a spike in g-forces to detect car accidents and auto-dial 911
Safety feature is also planned for Apple Watches expected in 2022, according to company documents
Context & Ripple Effects
Apple had already been building the Watch toward continuous personal monitoring, from tested sleep tracking to reported plans for blood-pressure and fertility-related sensors. Adding collision recognition broadens that roadmap from health measurement to emergency response across Apple’s two largest personal-device lines.
The later record shows the core trade-off in such a feature: iPhone 14 false alarms during rollercoaster rides and phone drops tied the feature’s usefulness to how well Apple distinguishes real crashes from high-force everyday events.
First-order effects
- Apple’s planned iPhone and Watch lineup gains an automated emergency-contact function, making sensor interpretation—not merely hardware inclusion—the immediate product requirement.
- Apple must tune the g-force trigger across two device categories because false emergency calls directly undermine the reliability of the safety promise.
Second-order effects
- Emergency dispatchers and Apple users bear the cost of overly sensitive detection, while Apple’s device teams face pressure to reduce false positives without missing actual collisions.
- The feature gives Apple a safety-oriented reason to position iPhone and Watch together, alongside the broader Watch health-sensor roadmap.
Third-order effects
- If Apple continues adding health and safety sensors, the Watch’s role shifts toward an always-worn monitoring endpoint while the iPhone supplies a complementary emergency interface.
- Machine-perception features in consumer devices will increasingly be judged by operational accuracy in consequential actions, not just by whether a sensor can detect an event.
The trend: Apple is extending its personal devices from passive health tracking toward sensor-driven interventions that act when users may be unable to do so.