Models & Inference

These endpoints control the AI lifecycle: training models on your custom data and running inference to generate risk scores.

Calibrate Model

Triggers the asynchronous training of an anomaly detection model using the files uploaded to a mission.

• Overview: Requires at least 15 files uploaded to the mission.
• Endpoint: /api/calibrate
• Method: POST

Request Headers

Key	Value	Required
Content-Type	application/json	Yes
X-API-KEY	Your-API-Key	Yes

Request Body

Parameter	Type	Required	Description
robot_id	uuid	Yes	The UUID of the robot.
mission_id	uuid	Yes	The UUID of the mission.
robot_type	string	Yes	The hardware type. Allowed: SPOT, UR6.
blob_names	list[str]	No	List of unique_filenames to train on. Defaults to all.
epochs1	int	No	Training epochs (Default: 100).
batch_size	int	No	Batch size for training (Default: 128).
retrain	boolean	No	Set true to force retraining (Default: false).

Success Response

Code: 200 OK

{
  "details": {
    "task_id": "a493b453-e29b-41d4-a716-446655440000"
  },
  "message": "Training task queued"
}

Possible Errors

• 400 Bad Request: Expected 15 files, but found X.
• 400 Bad Request: Model already exists.
• 401 Unauthorized: Invalid or missing API Key.

Sample Code (Python)

import requests

url = "[https://api.ridescan.ai/api/calibrate](https://api.ridescan.ai/api/calibrate)"
headers = {
    "Content-Type": "application/json",
    "X-API-KEY": "YOUR_API_KEY"
}
payload = {
    "robot_id": "e7201eff-e29b-41d4-a716-446655440000",
    "mission_id": "9384ac10-e29b-41d4-a716-446655440000",
    "robot_type": "SPOT",
    "epochs1": 150,
    "retrain": True
}

response = requests.post(url, json=payload, headers=headers)
print(response.json())

---

Process Files (Inference)

Runs the trained model against specific files to generate Risk Scores.

• Overview: The mission must have a trained model (calibration_completed) before you can run inference.
• Endpoint: /api/process
• Method: POST

Request Headers

Key	Value	Required
Content-Type	application/json	Yes
X-API-KEY	Your-API-Key	Yes

Request Body

Parameter	Type	Required	Description
robot_id	uuid	Yes	The UUID of the robot.
mission_id	uuid	Yes	The UUID of the mission.
blob_names	list[str]	No	List of unique_filenames to analyze. Defaults to all.
device	string	No	Device to run inference on: cpu or cuda (Default: cpu).

Success Response

Code: 200 OK

{
  "file sent": [
    "1a3a81da-e29b-41d4-a716-446655440000_Day16.bag",
    "ac8d4704-e29b-41d4-a716-446655440000_Day17.bag"
  ],
  "message": "Processing queued"
}

Possible Errors

• RS-MOD-001: Model not found (Training not completed).
• 401 Unauthorized: Invalid or missing API Key.

Sample Code (Python)

import requests

url = "[https://api.ridescan.ai/api/process](https://api.ridescan.ai/api/process)"
headers = {
    "Content-Type": "application/json",
    "X-API-KEY": "YOUR_API_KEY"
}
payload = {
    "robot_id": "e7201eff-e29b-41d4-a716-446655440000",
    "mission_id": "9384ac10-e29b-41d4-a716-446655440000",
    "device": "cpu"
}

response = requests.post(url, json=payload, headers=headers)
print(response.json())

---

Get Model & Mission Details

Retrieves the current status of the AI model and the aggregated risk scores.

• Endpoint: /api/model/mission-details
• Method: GET

Request Headers

Key	Value	Required
Content-Type	application/json	Yes
X-API-KEY	Your-API-Key	Yes

Query Parameters

Parameter	Type	Required	Description
mission_id	uuid	Yes	The Public UUID of the mission.

Success Response

Code: 200 OK

{
  "calibration_status": "calibration_completed",
  "epochs": 150,
  "files": [
    {
      "filename": "Day10.bag",
      "risk_score": 5.4
    }
  ],
  "inference_possible": true,
  "inference_status": "processing_completed",
  "mission_avg_risk_score": 12.5,
  "mission_length": 2073,
  "robot_type": "SPOT",
  "training_time": "Wed, 07 Jan 2026 10:30:00 GMT"
}

Possible Errors

• RS-MSN-002: Mission ID not found.
• 401 Unauthorized: Invalid or missing API Key.

Sample Code (Python)

import requests

url = "[https://api.ridescan.ai/api/model/mission-details](https://api.ridescan.ai/api/model/mission-details)"
headers = {
    "Content-Type": "application/json",
    "X-API-KEY": "YOUR_API_KEY"
}
params = {
    "mission_id": "9384ac10-e29b-41d4-a716-446655440000"
}

response = requests.get(url, params=params, headers=headers)
print(response.json())