17 IoT Systems Evolution: Computing and Placement
17.1 Start With the Story
Computing improved, but a sensor system still cannot send every byte to an unlimited cloud without cost, delay, or energy consequences. The team must decide where processing belongs and explain why distributed placement helps some workloads without becoming a universal answer.
17.2 Overview
This route connects computing economics to edge, fog, and cloud placement, then tests myths, pitfalls, technology groups, and architecture choices.
This is part 2 of 2. Review IoT Systems Evolution: Cycles and Convergence when you need the first route.
17.3 Learning Objectives
By the end of this chapter, you will be able to:
- explain the economics of distributed IoT computing
- choose edge, fog, and cloud placement from requirements
- test evolution claims against power and architecture evidence
17.4 Chapter Roadmap
Follow the original sections below in order. They begin at the reviewed split boundary and keep every worked example, figure, check, and supporting banner with the section that owns it.
17.5 Distributed Computing Economics
Earlier economics often favoured centralised compute
When single-core performance was still improving rapidly (30-40% per year), it made economic sense to:
- Build faster central servers for data processing
- Keep devices simple and dumb (just sensors sending data)
- Process everything in data centers with the fastest CPUs
Later economics widened compute-placement options
When single-core performance plateaued, chip manufacturers pivoted to:
- Multi-core processors: 2, 4, 8, 16+ cores on a single chip
- Specialized accelerators: GPUs for graphics, NPUs for AI
- Energy efficiency: Focus on performance-per-watt instead of raw speed
1. Cheap, Powerful Microcontrollers
| Microcontroller | Year | Price | Capabilities |
|---|---|---|---|
| Intel 8051 | 1980 | $5-10 | 8-bit, 12 MHz, no networking |
| ARM Cortex-M0 | 2009 | $0.50-1 | 32-bit, 48 MHz, low power |
| ESP32 | 2016 | $2-4 | Dual-core 240 MHz, Wi-Fi/Bluetooth, 520KB RAM |
| Raspberry Pi Pico | 2021 | $4 | Dual-core 133 MHz, 264KB RAM, extensive I/O |
Today’s $0.50 microcontroller has more computing power than a $10,000 desktop computer from 1995.
2. AI at the Edge
The proliferation of specialized neural network accelerators makes on-device AI feasible:
- Google Coral Edge TPU: 4 trillion operations/sec, $25, 2W power
- NVIDIA Jetson Nano: 472 GFLOPS, $99, 5-10W power
- Apple Neural Engine: 15.8 trillion ops/sec, integrated in phones
Real Impact: A security camera can now run face detection locally at 30 fps instead of streaming video to the cloud (99.9% bandwidth reduction).
3. Sensor Fusion on Device
Modern IoT devices can fuse data from multiple sensors in real-time:
- Smartphones: 24+ sensors (accelerometer, gyroscope, magnetometer, GPS, barometer, light, proximity, etc.)
- Drones: IMU (9-axis), GPS, ultrasonic, optical flow, cameras - all fused for stable flight
- Wearables: Heart rate, ECG, SpO2, temperature, accelerometer combined for health insights
4. Fog Computing Architecture
The abundance of cheap computing enables hierarchical processing where data is processed at the most appropriate layer:
- Device layer: Immediate response (sensor -> actuator in <10 ms)
- Edge layer: Local analytics (factory floor aggregation)
- Fog layer: Regional processing (city-wide traffic optimization)
- Cloud layer: Long-term storage and complex AI training
To test distributed computing economics, open the diagram in Figure 17.1. Fog Computing Hierarchy supplies one named condition; Less raw data supplies the necessary comparison for fog computing hierarchy showing where device, edge, fog, and cloud layers process data according to latency and coordination needs.
Compare Fog Computing Hierarchy with Less raw data inside the visual at Figure 17.1. Next find More raw data, which completes the scope of fog computing hierarchy showing where device, edge, fog, and cloud layers process data according to latency and coordination needs. The decision in distributed computing economics must preserve that labelled boundary.
Key insight: As data moves upward through the layers, volume decreases but insight value increases. A temperature sensor generates 10 readings per second (raw data), but the edge summarizes this into “temperature is trending up” (aggregated insight), and the cloud uses this pattern across thousands of sensors to predict equipment failures (strategic intelligence).
Checkpoint: Power Efficiency and Integration
You now know:
- Ideal Dennard scaling linked smaller dimensions with lower voltage and roughly controlled power density; real technologies increasingly departed from that ideal.
- Across the early-to-mid-2000s, leakage, power density, interconnect and thermal constraints slowed reliance on frequency growth alone.
- Multicore, efficiency and specialised logic — combined with continued integration and falling MCU and radio costs — widened practical IoT endpoint options.
17.6 IoT Systems Evolution Myths
Misconception 1: “IoT is just about connecting things to the internet.” Simple connected sensors are valid IoT endpoints. Local intelligence is an architecture choice used when latency, resilience, privacy, bandwidth or safety justifies processing near the physical process.
Misconception 2: “Moore’s Law ended, so computing improvement stopped.” Long-run transistor-density gains continued after real technologies departed from ideal Dennard scaling. Benefits increasingly appeared through multicore designs, specialised accelerators, integration and energy efficiency rather than frequency growth alone.
Misconception 3: “Cloud computing and IoT are competing approaches.” Cloud and edge are complementary, not competing. Modern IoT architectures use both: edge devices handle time-sensitive sensing and local actuation, while cloud services handle long-term storage, model training and fleet coordination. The mid-2000s power-efficiency pivot was one contributor to wider placement options; integration, component costs, networks, software platforms and application demand also mattered.
Misconception 4: “IoT just needed cheaper hardware to happen.” Lower component cost was one contributor, not a prerequisite acting alone. Practical deployments also depended on suitable sensors and radios, network coverage, protocols, software platforms, device management, security, integration effort and application demand.
17.7 Systems Evolution Pitfalls
Pitfall 1: Assuming “newer era = abandon previous era” Each computing era builds on top of previous ones rather than replacing them. Companies that rip out centralized servers in favor of edge-only architectures discover they still need cloud for ML training, fleet management, and long-term analytics. The correct approach is a complementary multi-tier architecture, not wholesale replacement.
Pitfall 2: Extrapolating Moore’s Law linearly into device cost A $0.50 microcontroller does not mean a $0.50 IoT device. The bill of materials (BOM) for a complete IoT node includes the MCU ($0.50-$4), wireless radio ($1-$5), antenna ($0.20-$2), power regulation ($0.50-$2), sensors ($0.50-$10), PCB and assembly ($1-$5), and enclosure ($1-$10). Total device cost is typically 10-50x the MCU cost alone. Budget accordingly.
Pitfall 3: Ignoring the “last mile” of the 10x cycle The 10x pattern shows device counts growing from billions to trillions, but the last trillion devices are the hardest. They require sub-$1 hardware, 10+ year battery life, and operation in harsh environments (underwater, underground, extreme temperatures). Do not assume the next 10x will arrive at the same pace as previous cycles — physics and economics impose harder constraints at each step.
Pitfall 4: Confusing clock speed with real-world performance A 240 MHz ESP32 is not “60x slower” than a 14 GHz desktop CPU in practical IoT tasks. Modern microcontrollers have hardware peripherals (DMA, hardware crypto, radio baseband) that offload work from the CPU. For sensor reading, protocol handling, and local inference, a $4 MCU often matches or outperforms what a $500 desktop achieves because the workload is I/O-bound, not compute-bound.
17.8 Computing Power Revolution
Impact on IoT Economics:
| Metric | 2005 | 2025 | Change |
|---|---|---|---|
| Microcontroller cost | $5-10 | $0.50-2 | 10x cheaper |
| Microcontroller power | 100 mW typical | 5 mW sleep, 50 mW active | 20x more efficient |
| Wireless module cost | $50-100 | $2-5 | 20x cheaper |
| Cloud storage cost | $1/GB/month | $0.01/GB/month | 100x cheaper |
| Sensor cost | $10-50 | $0.20-2 | 25x cheaper |
Result: The total cost of an IoT device dropped from ~$200+ (2005) to ~$5-20 (2025), making it economical to connect billions of everyday objects.
Use Figure 17.2 to prepare the decision in computing power revolution. The diagram names Systems Evolution: Economics and Cost, the two anchors needed to assess systems evolution economics: falling hardware, connectivity, and cloud costs lowered the adoption threshold for iot.
At Systems Evolution: Economics in Figure 17.2, compare the diagram with Cost; then locate unit. That labelled check bounds systems evolution economics: falling hardware, connectivity, and cloud costs lowered the adoption threshold for iot. For computing power revolution, retain unit as evidence for the resulting choice.
17.9 How Distributed Placement Can Help
Real-World Example: Smart Factory Comparison
Illustrative centralized baseline:
- 1,000 sensors -> Central server processes all data -> Actuators respond
- Latency: 100-500 ms (sensor -> cloud -> actuator)
- Bandwidth: 1 Gbps fiber needed for real-time data
- Single point of failure: Server down = factory stops
- Cost: $100K server + $50K/year bandwidth
Illustrative distributed alternative:
- 1,000 smart sensors with $2 microcontrollers -> Local edge processing -> Immediate actuation
- Latency: 1-10 ms (sensor -> edge -> actuator)
- Bandwidth: 10 Mbps (only aggregate data to cloud)
- Resilient: Local loops continue if cloud fails
- Cost: $2K in microcontrollers + $1K/year bandwidth
Savings: $98K upfront + $49K/year by distributing computing to edge devices
Checkpoint: Distributed Placement
You now know:
- The factory example compares 1,000 sensors through a $100K central server against 1,000 smart sensors using $2 microcontrollers.
- The chapter separates upfront savings from operating savings: $100K minus $2K gives the $98K edge swing, while bandwidth falls from $50K/year to $1K/year.
- Fog hierarchy is a placement model: device for immediate work, edge for local analytics, fog for regional coordination, and cloud for long-term storage and training.
17.10 Technology Groups Supporting Modern IoT
1. Abundant Cheap Computing
- Moore’s Law continues (transistor count still doubling)
- Focus shifted from speed to efficiency and specialization
- $0.50 microcontrollers rival 1990s desktops
2. Energy Efficiency Revolution
- Performance-per-watt improved 100x from 2005-2025
- Battery-powered devices can run for years
- Energy harvesting becomes viable for some applications
3. Specialized Accelerators
- Neural network accelerators: AI on $25 chips
- Cryptographic engines: Secure communication with minimal overhead
- Radio modules: Wi-Fi/Bluetooth/LoRa on single chip
Result: Continued integration, lower-cost components, improved efficiency, networking, software platforms and application demand jointly widened the range of practical connected products.
17.11 Why Large-Scale IoT Reached Practical Scale Gradually
Several trends converged over time:
- Earlier deployments: Distributed nodes often faced higher component, power, integration and operating costs, so central processing remained attractive for many workloads.
- Mid-2000s onward: Power and thermal constraints increased emphasis on efficiency, multicore processors and specialised logic.
- 2010s: Continued integration, lower-cost MCUs, radios and sensors, broader networking, cloud and device platforms, and maturing protocols reduced deployment friction.
- Later deployments: Edge accelerators and improved tooling widened the workloads that could run near physical processes.
The mid-2000s power-efficiency pivot was one contributor. Continued integration, lower-cost microcontrollers, radios and sensors, networking, software platforms and application demand jointly widened practical IoT and edge-computing options.
17.12 Connecting to IoT Architecture
These converging technology and deployment trends support the architectural patterns studied later in this module:
- Edge Computing: Places work near devices when latency, resilience, privacy or bandwidth justifies it
- Fog Computing: Hierarchical processing requires powerful intermediate nodes
- Edge AI/ML: Requires specialized neural accelerators now available in $25 chips
- Wireless Sensor Networks: Nodes need computation for routing, data fusion, and coordination
No single event produced these patterns. Hardware integration and efficiency, component cost, networks, software platforms and application demand converged to widen useful compute-placement choices.
To test connecting to iot architecture, open the diagram in Figure 17.3. How Technology Convergence Supports IoT Patterns supplies one named condition; Hardware, networks, software, and demand jointly widened supplies the necessary comparison for multi-factor convergence widened practical options for edge computing, fog computing, edge ai and wireless sensor networks.
Compare How Technology Convergence Supports IoT Patterns with Hardware, networks, software, and demand jointly widened inside the visual at Figure 17.3. Next find Multi-factor, which completes the scope of multi-factor convergence widened practical options for edge computing, fog computing, edge ai and wireless sensor networks. The decision in connecting to iot architecture must preserve that labelled boundary.
We just turned the history into architecture enablers. The checks and cases that follow ask you to choose placement by constraint, not by slogan.
17.13 Computing Evolution Check
Question 4: Which challenge contributed to the departure from ideal Dennard scaling in the early-to-mid-2000s?
a) Transistors became too expensive to manufacture b) Software couldn’t keep up with hardware improvements c) Transistors became so small that leakage current created excessive heat d) Global chip manufacturing capacity was reached
17.13.1 Answer
c) Transistors became so small that leakage current created excessive heat — As process technologies advanced across multiple generations, supply-voltage reduction slowed while leakage, interconnect, power-density and cooling constraints became harder to manage. These interacting limits weakened the ideal constant-field relationship and reduced the practicality of relying on frequency growth alone; there was no single universal 65 nm switch.
Question 5: A factory compares centralized vs distributed IoT architectures. The centralized approach uses a $100K server with 100-500ms latency. The distributed approach uses 1,000 smart sensors with $2 microcontrollers and 1-10 ms latency. What is the approximate upfront cost savings of the distributed approach?
a) $2,000 b) $50,000 c) $98,000 d) $148,000
17.13.2 Answer
c) $98,000 — The distributed approach costs approximately $2,000 for microcontrollers (1,000 x $2) versus $100,000 for the centralized server. The upfront saving is $100K - $2K = $98K. Separately, bandwidth falls from $50K/year to $1K/year, an ongoing saving of $49K/year.
Question 6: Why were large-scale distributed IoT deployments less common in the 1990s than today?
a) There was no consumer demand for connected devices b) Wireless networking standards had not yet been developed c) Sensors, compute, radios, networks, platforms and deployment economics had not yet converged at modern scale d) Governments restricted IoT technology for military use
17.13.3 Answer
c) Sensors, compute, radios, networks, platforms and deployment economics had not yet converged at modern scale — Earlier connected systems existed, but large fleets were harder to build and operate. Later integration, efficiency, component-cost, networking, software-platform and demand trends jointly widened the practical deployment space.
Question 7: Which statement best relates the mid-2000s power-efficiency pivot to cheaper computing, energy efficiency and specialised accelerators?
a) They were all funded by the same government research program b) The pivot was one contributor, while integration, component costs, software, networking and demand also mattered c) They were developed by the same semiconductor company d) They were all required by smartphone manufacturers
17.13.4 Answer
b) The pivot was one contributor, while integration, component costs, software, networking and demand also mattered — Power and thermal constraints increased emphasis on performance per watt, multicore designs and specialised logic. Continued integration, lower-cost microcontrollers, radios and sensors, networking, software platforms and application demand jointly widened practical IoT options.
17.14 Interactive Quiz: Match Concepts
17.15 Interactive Quiz: Sequence the Steps
17.16 Label the Diagram
17.17 Code Challenge
17.18 Summary
17.18.1 Key Concepts
In this chapter, you learned how several decades of computing, networking and deployment trends helped make large-scale IoT practical:
- Technology cycles follow a 10x pattern — each era brings approximately 1,000x more devices at 1/100th the cost, driven by the dual forces of lower prices and improved functionality
- Five phases of connectivity evolved from connecting computers (1969) to connecting people (1991-2012) to connecting things (2010+), with each phase building on the infrastructure of the previous one
- Real technologies departed from ideal Dennard scaling as voltage, leakage, interconnect, power-density and thermal constraints reduced reliance on frequency growth alone.
- The mid-2000s power-efficiency pivot was one contributor to greater emphasis on efficiency, multicore processors and specialised logic.
- Integration and component-cost trends expanded the capability available in microcontrollers, radios and sensors, but complete-node and deployment costs still matter.
- Networks, software platforms and application demand converged with hardware progress to widen practical IoT options.
- Edge computing is a placement choice used when latency, resilience, privacy, safety or bandwidth makes local processing valuable; cloud and edge remain complementary.
17.18.2 Looking Back, Looking Forward
The evolution from mainframes to IoT is not just a story of shrinking hardware — it is a story of expanding capability and deployment choice. Integration, efficiency, radios, networks and software made connected computing practical for many more physical processes. Understanding complete-system economics is essential for deciding where to process data, how much intelligence to put at the edge and when an IoT deployment creates enough value to justify its lifecycle cost.
17.19 Knowledge Check
17.20 Quiz: Systems Evolution to IoT
17.21 Edge vs Cloud IoT Processing
Hardware integration and efficiency, lower component costs, networking and software platforms widened edge-processing options, but placement remains a workload decision. Use this framework to decide where intelligence belongs.
17.22 The Decision Matrix
Use these placement questions as a compact decision matrix:
- Latency requirement: process at the edge when the action must complete in under 100 ms; cloud is reasonable when a network round trip of roughly 50-200 ms is acceptable.
- Data volume: process high-volume streams, such as video, near the source; send compact events or summaries to cloud services.
- Connectivity reliability: keep safety and continuity logic local when links are intermittent; use cloud processing when the connection is reliable.
- Power budget: let sleepy edge devices duty-cycle when the radio is expensive; use cloud processing when the device can afford continuous communication.
- Processing complexity: run simple inference and rules locally; reserve training, multi-model analysis, and heavy storage for cloud services.
- Security and privacy: keep sensitive biometric, health, or operational data local where possible; transmit only the derived evidence the workflow needs.
- Fleet learning: use cloud aggregation when cross-site learning matters; keep isolated fast-response logic at the edge.
- Device cost: choose edge AI only when the product can afford the $10-$50 compute path; use a cheaper MCU plus radio when thin-edge telemetry is enough.
17.23 Four Common Architecture Patterns
Pattern 1: Edge-Only (No Cloud)
- Example: Industrial safety system cutting power to machinery when hazard detected
- Why: 10 ms latency requirement, must work during network outages, safety-critical
- Edge compute: $25 microcontroller with local ML model
- Cloud role: None (fully autonomous)
Pattern 2: Cloud-Only (Thin Edge)
- Example: Smart parking sensor transmitting occupancy state every 30 seconds
- Why: Simple binary data (occupied/vacant), no latency requirement, battery-powered
- Edge compute: $2 MCU + LoRaWAN radio
- Cloud role: All analytics (occupancy patterns, pricing optimization, reporting)
Pattern 3: Hybrid (Edge Inference + Cloud Training)
- Example: Security camera with edge face detection
- Why: Bandwidth savings (99% reduction vs. streaming raw video), privacy (faces processed locally)
- Edge compute: $35 processor with neural accelerator runs inference at 30 fps
- Cloud role: Trains updated face detection models weekly using edge-collected metadata (not faces)
Pattern 4: Tiered (Device → Edge Gateway → Cloud)
- Example: Factory with 500 sensors → 5 edge gateways → Cloud
- Why: Sensors are low-cost ($2 each), gateway aggregates/filters data from 100 sensors, cloud provides fleet analytics
- Edge compute: $200 gateway with 16-core processor handles real-time aggregation
- Cloud role: Long-term storage, cross-factory optimization, predictive models
17.24 Smart Agriculture Decision
Scenario: 500-acre farm needs soil moisture monitoring to optimize irrigation.
Option A: Cloud Processing
- Sensor: $12 (MCU + LoRaWAN radio + moisture probe)
- Sends reading every 15 minutes to cloud
- Cloud analyzes all 50 sensors, sends irrigation commands
- Latency: 5-30 seconds
- Cost per sensor: $12 hardware + $3/year connectivity = $15/year
- 50-sensor network: $750 hardware + $150/year OpEx
Option B: Edge Gateway Processing
- Sensor: $8 (simple analog probe + LoRaWAN)
- Edge gateway: $350 (solar-powered, runs local irrigation optimization model)
- Gateway reads 50 sensors every 5 minutes, controls irrigation valves locally
- Cloud used only for historical storage and model updates
- Latency: <1 second (local control loop)
- Cost: $400 sensors + $350 gateway + $30/year connectivity = $780/year
Option C: Hybrid Approach
- 45 low-cost sensors ($8 each): Report to cloud every 30 min
- 5 reference sensors ($50 each): Edge processing for critical zones with 1-minute local control
- Cloud analyzes 45-sensor trends, local edge handles fast response in critical areas
- Cost: $360 + $250 = $610 hardware + $75/year OpEx
Decision Factors:
- Irrigation response time: Crops can tolerate 5-30 second delays (cloud OK), except in sandy soil zones (need edge)
- Connectivity: LoRaWAN coverage good but not perfect (edge provides autonomy)
- Learning: Cloud aggregates multi-field patterns for better predictions (cloud advantage)
Optimal Choice: Option C (Hybrid)
- Combines low-cost broad coverage (cloud) with fast local control where needed (edge)
- 25% lower cost than full edge (Option B)
- Better responsiveness than pure cloud (Option A) for critical zones
17.25 Quick Decision Algorithm
IF latency < 100ms OR connectivity unreliable OR data volume > 1 MB/sec
→ Edge processing required
ELSE IF processing complexity high OR need fleet learning
→ Cloud processing preferred
ELSE IF budget tight AND latency tolerance >5 sec
→ Cloud processing (thin edge)
ELSE
→ Hybrid (edge inference + cloud training)17.26 Cost Curves: When Edge Becomes Cheaper
Use this cost curves: when edge becomes cheaper section as a guided decision record, not as a list to memorise. First identify the stated input, assumption, or scenario; then compare each option on the same units and time boundary. Next check which value changes the outcome and which evidence would reveal an invalid assumption. For cost curves: when edge becomes cheaper, the useful result is the reasoning chain: observed condition, governing constraint, calculation or classification, and operational consequence. Record that chain before choosing an answer or carrying a value into the next section. Where the panel supplies several choices, reject each distractor against the chapter’s named mechanism instead of relying on wording cues. Where it supplies a table or timeline, compare rows at like-for-like scale and preserve the difference between an early indication, an actionable threshold, and a final outcome. This turns cost curves: when edge becomes cheaper into evidence that can be reviewed, recalculated, and connected to the running design narrative.
The breakeven point where edge processing costs less than cloud:
| Data Volume | Breakeven Device Count | Explanation |
|---|---|---|
| 10 kB/sec | 1 device | Even 1 device streaming 10 kB/sec costs $15/month cellular; $50 edge chip amortizes in 3 months |
| 1 kB/sec | 10 devices | $1.50/month cellular × 10 = $15/month; edge gateway ($200) pays back in 13 months |
| 100 bytes/sec | 100 devices | LPWAN costs $3/year/device, cloud competitive; edge only if latency matters |
Key insight: Technology convergence widened edge-processing options, but the appropriate architecture depends on latency, data volume, connectivity, power, privacy, safety and fleet-learning constraints. There is no universal “always edge” or “always cloud” answer.
Checkpoint: Edge or Cloud
You now know:
- Put logic at the edge when latency is under 100 ms, connectivity is unreliable, or high-volume data should be reduced near the source.
- Keep cloud work for training, long-term storage, cross-site learning, and fleet analytics where 50-200 ms network timing is acceptable.
- The smart-agriculture case chooses hybrid placement because broad sensing, critical-zone response, cloud learning, and cost all pull in different directions.
17.27 Concept Check
17.28 Concept Relationships
- 10x technology cycles connect to IoT Introduction: each era, from mainframe to PC to mobile to IoT, made many more devices economically viable at much lower unit cost.
- The mid-2000s power-efficiency pivot connects to Edge Computing as one contributor to more efficient processing; integration, lower-cost components, networks, software and application demand jointly widened placement options.
- Moore’s Law connects to hardware platforms: transistor-density gains still matter because they let low-cost microcontrollers include more memory, radios, security blocks, and local processing.
- Distributed versus centralized computing connects to fog computing: modern IoT systems use tiered edge, fog, and cloud placement rather than a single compute location.
- The energy-efficiency revolution connects to power management: performance-per-watt improvements make battery-powered sensing and decade-scale field deployments practical.
17.29 See Also
Related Evolution Chapters:
- Device Evolution - How Embedded → Connected → IoT progression parallels computing evolution
- IoT History - Paradigm shifts and lessons from technology transitions
- Industrial IoT Fundamentals - Fourth Industrial Revolution context for computing economics
Architecture Enabled by Evolution:
- Edge Computing - Local processing made viable by cheap microcontrollers
- Fog Computing - Hierarchical processing across edge-fog-cloud tiers
- Edge AI/ML - On-device inference using specialized accelerators
Technology Deep Dives:
- Microcontrollers - ARM Cortex-M series and modern IoT chips
- Wireless Technologies - How wireless module costs dropped 20x since 2005
- Power Management - Ultra-low-power design for 10-year deployments
17.30 What’s Next
| Direction | Chapter | Description |
|---|---|---|
| Next | IoT Perspectives | Stakeholder definitions and viewpoints after the systems-evolution view |
| Previous | IoT History | Paradigm shifts and lessons from technology transitions |
| Related | Edge Computing | Local processing made viable by cheap microcontrollers |
| Related | Device Evolution | Embedded to Connected to IoT progression |
| Hub | Quiz Navigator | Test your understanding across all chapters |