How Does AI in Smart Socks Work?
The AI engine in Mi Terro Care Socks performs three core functions: personal baseline learning, anomaly detection, and multi-signal correlation. Together, these detect foot complications 24-72 hours before symptoms — catching problems invisible to both the patient and periodic clinical exams.
Personal Baseline Learning
Every person's feet have unique thermal and biomechanical signatures. The AI establishes a personal baseline during the first 7 days of use, collecting data from all 11 sensors across different activities and times of day. It identifies:
- Normal temperature range per zone
- Typical gait pattern and variability
- Standard pressure distribution during standing and walking
- Circadian temperature rhythms
After calibration, the system monitors for deviations from your personal baseline — not population averages. This makes detection far more sensitive and specific.
Anomaly Detection
The AI continuously compares real-time data against expected values, detecting:
- Sustained temperature asymmetry — same zone on left/right foot differs by 2.2°C+ for an extended period
- Rate-of-change spikes — temperature rises faster than baseline predicts
- Gait deterioration — stride, cadence, or symmetry decline beyond personal range
- Pressure redistribution — weight shifts to unusual zones
Multi-Signal Correlation
The most powerful capability: cross-correlating signals across all three sensing modalities simultaneously.
A temperature spike alone might be normal variation. But a temperature spike at the metatarsal combined with increased pressure at the same zone and decreased stride length creates a high-confidence alert for tissue injury.
In beta testing, 12 anomalies flagged by the AI correlated with self-reported inflammation events — early clinical validation.
Edge Computing and Privacy
AI runs on the wall-plug hub using edge computing — data is processed locally before cloud transmission. Benefits:
- Faster alerts (no cloud latency)
- Monitoring continues during internet outages
- Enhanced data privacy
Frequently Asked Questions
How long does the AI take to learn my baseline?
7 days. After calibration, the baseline continues updating incrementally as the system learns long-term patterns.
Does the AI replace my doctor?
No. The AI supplements clinical care by providing objective, continuous data between clinic visits. When it detects an anomaly, it alerts patients, family, and care team — who make clinical decisions.