How iOS 15 Walking Steadiness Tracks Cadence

Apple’s iOS 15 introduced the Walking Steadiness feature to assess fall risk by analyzing everyday mobility directly through the iPhone. This article explains how the system monitors walking cadence over time, detailing the sensors involved, the algorithmic processes used to establish baseline patterns, and how subtle shifts in step frequency are flagged to help users understand their long-term stability and health.

Built-in Sensor Data Collection

Walking steadiness relies on the motion sensors built directly into the iPhone, primarily the three-axis accelerometer and the three-axis gyroscope. When you carry your iPhone in a pocket or secured near your waist during routine daily movement, these sensors register three-dimensional acceleration, tilt, and rotational forces generated by your hips and legs.

Cadence, defined as the number of steps taken per minute, is measured by capturing the rhythmic vertical and forward acceleration spikes that occur during each heel strike. Rather than requiring continuous active sessions, iOS 15 passively captures these bursts of movement whenever you walk on flat, even ground for a sustained period.

Algorithmic Processing and Noise Filtering

Raw sensor data is inherently noisy due to variable phone positioning, sitting down, or brief, irregular movements. iOS 15 utilizes custom machine-learning models and signal-processing algorithms to isolate genuine, steady-state walking from other physical activities:

  1. Gait Isolation: The system validates that the user is moving in a consistent forward trajectory rather than shifting positions in a chair or taking short, disrupted steps.
  2. Step Detection: Once a steady gait is confirmed, the algorithm identifies individual step intervals with high precision, converting the time between successive steps into an instantaneous cadence measurement.
  3. Contextual Filtering: Periods of running, ascending stairs, or carrying the phone in a swinging hand are filtered out to maintain data integrity.

Long-Term Tracking and Baseline Comparison

To detect meaningful changes, iOS 15 aggregates these individual cadence samples over days, weeks, and months within the Health app:

  • Baseline Establishment: During the initial weeks of use, the operating system establishes a personalized baseline for your typical cadence along with other metrics, such as walking speed, step length, and double support time.
  • Trend Detection: iOS 15 applies statistical smoothing across historical data to identify long-term trajectories. A temporary dip due to minor fatigue is treated as normal variation, whereas a sustained decrease in step frequency over consecutive weeks is registered as a downward trend.
  • Steadiness Classification: Cadence data is combined with walking asymmetry and step length to categorize your overall stability into three levels: OK, Low, or Very Low.

If cadence significantly decreases or becomes markedly erratic over time, the Health app logs this shift as part of your overall steadiness profile and can alert you to a rising risk of falling, prompting lifestyle adjustments or medical consultation.