Analytics & ML · Study deck
Edge Acquisition: Control Paths and Bandwidth
A cloud command and an edge alarm do not share the same safe delay.
Data Dora is your guide for this deck.
After studying this chapter
Learning objectives
You will be able to:
- Explain: The widening orders-of-magnitude gap means one forwarding policy cannot fit every source: sparse events may travel intact, periodic measurements may be batched, and continuous media often needs local selection or inference.
- Explain: This connects sample timing to the chapter's running design decision: faster acquisition must be justified by signal bandwidth or decision latency, then paid for with storage, radio energy, or local reduction.
- Explain: Each tenfold increase in frequency produces ten times as many 20-byte records, so the visual growth is arithmetic rather than a property of a particular sensor.
Major section
Cloud and Edge Control Paths
The: DESIGN INFERENCE explains that central history is simple, but a WAN failure interrupts the decision unless a safe fallback exists.
- That split improves continuity but makes gateway update and replay ownership explicit.
- The latency chips and continuity ledger show why urgency, sensitivity, and replay needs—not data volume alone—set placement.
Major section
Data Generation Patterns
Understanding data generation patterns is essential for designing efficient edge acquisition systems.
- Different device types produce vastly different data volumes and require different handling strategies.
- This carries the category model into the running acquisition narrative by turning device capability into a bandwidth, storage, and processing obligation.
Major section
Mobile figure summary
Big Things: Servers and PLCs have full IP connectivity, so they can upload structured data directly to the cloud.
- Small IP Things: Smart cameras and other embedded devices connect over Wi-Fi or cellular and can pre-filter data before upload.
Major section
Mobile figure summary (continued)
Non-IP Things: Simple sensors use Zigbee, BLE, Modbus, or similar buses and rely on a nearby gateway for translation and aggregation.
- Connectivity path: Non-IP sensor -> gateway -> cloud, while Big Things and Small IP Things can usually reach the cloud without protocol translation.
- Each tenfold increase in frequency produces ten times as many 20-byte records, so the visual growth is arithmetic rather than a property of a particular sensor.
- This connects sample timing to the chapter's running design decision: faster acquisition must be justified by signal bandwidth or decision latency, then paid for with storage, radio energy, or local reduction.
Major section
Data Volume by Device Type
This view shows how data generation rates vary dramatically by device type, driving different edge processing strategies.
- The widening orders-of-magnitude gap means one forwarding policy cannot fit every source: sparse events may travel intact, periodic measurements may be batched, and continuous media often needs local selection or inference.
Major section
Inertial Measurement Example
High-frequency sensors like accelerometers and gyroscopes demonstrate why edge aggregation is critical.
- Accelerometer changes combine motion and gravity; gyroscope changes record rotation.
- A motion event can therefore span several axes and both sensor types, which is why a single average may erase useful structure.
Major section
Mobile figure summary
Edge aggregation reduces this to statistical summaries at 1 Hz.
- Accelerometer axes: X, Y, and Z linear acceleration are tracked together to capture movement and gravity.
- Edge takeaway: High-frequency six-axis data is useful locally, but the transmission layer should send aggregated summaries instead of every raw sample.
- Transmitting raw data as 16-bit integers would require ~1.2 KB/s -- unsustainable for battery-powered devices on LPWAN networks.
Deck summary
Key takeaways
The: DESIGN INFERENCE explains that central history is simple, but a WAN failure interrupts the decision unless a safe fallback exists.
- Understanding data generation patterns is essential for designing efficient edge acquisition systems.
- Big Things: Servers and PLCs have full IP connectivity, so they can upload structured data directly to the cloud.
- Non-IP Things: Simple sensors use Zigbee, BLE, Modbus, or similar buses and rely on a nearby gateway for translation and aggregation.
- This view shows how data generation rates vary dramatically by device type, driving different edge processing strategies.
Retrieval practice
Recall check 1 of 2
Data Dora says: answer from memory, then check your reasoning.
Q1A cold-store WAN link fails while the compressor still needs control decisions. Which design preserves the local decision path?
Show answer
Answer: C The edge path connects gateway decisions to a local actuator while cloud analytics receive evidence separately.
Retrieval practice
Recall check 2 of 2
Data Dora says: answer from memory, then check your reasoning.
Q2A gateway serves continuous high-volume sources and sparse event-driven sensors. What should guide its bandwidth plan?
Show answer
Answer: B The device-class comparison connects these properties to bandwidth, storage, and processing needs.
Print reference
Answers
Answer key.
- C · The edge path connects gateway decisions to a local actuator while cloud analytics receive evidence separately.
- B · The device-class comparison connects these properties to bandwidth, storage, and processing needs.