Diagnostics

Diagnose sensor faults

Send sensor readings to Cadreen. It checks for motor overcurrent, temperature warnings, bearing wear, and battery drain — before they become problems.

What can Cadreen diagnose?

Motor overcurrent, temperature warnings, bearing wear, and battery drain. Cadreen checks readings against known fault patterns first. If nothing's clear, it asks an AI model for a second opinion.

Normal sensor ranges

Variable	Default	Description
Motor current	5-10A	Normal operating range for most motors
Temperature	20-60C	Ambient and motor temperature
Vibration	0-2 mm/s	Normal vibration levels
Battery	30-100%	Healthy battery range

What faults look like

warning

Motor current: 15A

Motor drawing 15A when normal is 8A — too much current

warning

Temperature: 85C

Temperature at 85C — running hot, needs attention

warning

Vibration: 6 mm/s

Elevated vibration — possible bearing wear

critical

Battery: 8%

Battery critically low — device needs charging now

How does Cadreen catch safety delays?

Cadreen monitors every safety check in real time. If a safety check takes too long — whether because of network lag or a system delay — Cadreen automatically stops all devices. This happens in milliseconds, faster than any human could react.

Cadreen responds within 5 milliseconds on local devices, 50ms in the cloud
Checks continuously, multiple times per second
If a safety check is delayed, all devices are stopped immediately
If a device can't be stopped, it's marked unsafe and you're alerted

What health information does Cadreen track?

Every device reports its health — battery, temperature, CPU usage, and network speed. Cadreen uses this information to assign tasks wisely: devices with low battery are rested, and overloaded devices get fewer jobs.

Battery: 0.0-1.0 (percentage)
Temperature: Celsius
CPU/Memory: 0.0-1.0 (utilization)
Network latency: milliseconds

How does Cadreen store sensor data?

Cadreen stores sensor data so you can look back at what happened. On local devices, it keeps the most recent 1,000 readings. In the cloud, it stores data longer. Cadreen also extracts patterns from raw data — like 'motor is running hot' — and adds them to its knowledge.

Local storage: 1,000 most recent readings per device
Cloud storage: longer retention, query by device or time range
Automatic pattern detection from raw sensor streams
Old data is purged to keep storage efficient

How do I run diagnostics?

POST/api/v1/cadreen/devices/diagnose

Send sensor readings and get back fault detection. Cadreen checks each reading against its diagnostic rules.

Response

{
  "diagnoses": [
    {
      "fault_id": "f_abc123",
      "fault_type": "motor_overcurrent",
      "severity": "warning",
      "confidence": 0.92,
      "description": "Motor drawing 15.2A when normal range is 5-10A",
      "root_cause": "Possible mechanical jam or bearing failure",
      "remediation": "Inspect motor load and bearings. Check for obstructions.",
      "rule_id": "rule_motor_current"
    }
  ],
  "total": 1
}

Example

curl -X POST http://localhost:8080/api/v1/cadreen/devices/diagnose \
  -H "Content-Type: application/json" \
  -d '{
    "readings": [
      {"name": "motor_current", "value": 15.2, "unit": "A"},
      {"name": "temperature", "value": 82.0, "unit": "C"}
    ]
  }'

GET/api/v1/cadreen/devices/diagnostics/capabilities

See what Cadreen can diagnose. Returns the number of diagnostic rules loaded and devices tracked.

Response

{
  "diagnostic_rules": 12,
  "devices": 3,
  "message": "System can diagnose sensor faults and track devices"
}

Example

curl http://localhost:8080/api/v1/cadreen/devices/diagnostics/capabilities