Data Ingestion Endpoint

[CyberSamurai0]

Summary

A key project requirement is that our data is ingested into a cloud IoT platform for analysis and response.

Ideally, only the handheld module (with "controlled" software) is trusted to send telemetry, and handheld devices are not trusted/required to report their own position. However, the gateway is a late-stage, optional portion of the project. We need telemetry testing earlier, so we'll use a less secure handheld-to-cloud model for now.

Specification

The handheld device shall report its position to the cloud endpoint immediately after calculation.

• The handheld must (for now) possess cellular or Wi-Fi networking capabilities
• The handheld shall transmit its updated position to a REST API endpoint hosted by the cloud platform
• The handheld shall include a device-unique identifier alongside each position payload such that position can be correlated over time
• Request payloads must be sent as application/json and include a verified JWT for authentication/authorization.

The cloud platform will process the new data, act upon it, and store it for later.

• In early testing, historical data is not valuable. We might as well only retain the most recently transmitted position until we have a reliable positioning system.
• We should also set a maximum retention to keep storage costs down.
  • Also, perhaps we can find a way to summarize older data so we don't keep one event per second from three weeks ago...
• More details on analytics and actuation to come!

[CyberSamurai0]

@legacv I've designed some preliminary Python code in anticipation of your specs, lmk if we need adjustments

ble-positioning/handheld/azure_api.py

Lines 14 to 26 in a1d85aa

	headers = {
	"Content-Type": "application/json",
	"Authorization": f"Bearer {os.getenv('AZURE_AUTH_TOKEN')}", # Load this to memory only when we need it
	"X-CNIT-546": True,
	}
	# Build Body
	body = {
	"loc_north": x,
	"loc_east": y,
	"building": building_id,
	"floor": floor
	}