I85PREP
⌘ K
CIoTP study materials
CIoTP study resources

CIoTP practice questions

One question at a time. Build your CIoTP understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Planning IoT systems

When planning the network topology for a large IoT deployment, which architectural model most directly supports low-latency local decision-making while still forwarding summarized data to the cloud?

  1. A star topology with no local processing and a single centralized broker
  2. A purely cloud-centric model where every sensor reading is sent to a remote data center
  3. A mesh topology that only performs store-and-forward relaying without computation
  4. A fog or edge computing model that processes data near the devices before forwarding aggregates Correct answer.

Explanation

Fog/edge computing places processing capability close to the devices, enabling local low-latency decisions while sending only summarized or filtered data upstream. Pure cloud models add round-trip latency for every reading, and topologies alone do not define where computation occurs.

Question 2Planning IoT systems

A planner must choose a wireless technology for battery-powered soil moisture sensors spread across a multi-hectare farm, reporting a few times per day and needing multi-kilometer range. Which technology best fits these constraints?

  1. LoRaWAN Correct answer.
  2. 5G Ultra-Reliable Low-Latency Communication
  3. Zigbee
  4. Bluetooth Low Energy

Explanation

LoRaWAN is designed for low-power, low-data-rate, long-range communication, matching sparse agricultural reporting over kilometers. BLE and Zigbee have short range, and URLLC targets high-reliability low-latency use cases with higher power budgets.

Question 3Planning IoT systems

Which statement best describes the role of a message broker in an IoT data architecture plan?

  1. It decrypts all device traffic before forwarding it to storage
  2. It guarantees that every sensor message is processed exactly once end to end
  3. It replaces the need for any device identity or authentication mechanism
  4. It decouples publishers from subscribers by routing messages based on topics or queues Correct answer.

Explanation

A broker implements publish/subscribe semantics, decoupling data producers from consumers via topic or queue routing. It does not inherently guarantee end-to-end exactly-once processing, and authentication remains a separate concern.

Question 4Planning IoT systems

During planning, a team defines a requirement that device firmware must be updatable remotely across thousands of units. Which planning consideration is most important to address first?

  1. Maximizing the message payload size for telemetry
  2. Choosing a proprietary vendor-specific update tool without interoperability review
  3. Establishing a secure, authenticated over-the-air update pipeline with rollback capability Correct answer.
  4. Selecting the color scheme for the device management dashboard

Explanation

Remote firmware updates are a high-risk operation, so the plan must prioritize authenticated delivery and rollback to avoid bricking fleets or enabling malicious firmware. Payload size and UI choices are secondary to update integrity.

Question 5Planning IoT systems

A planner is sizing the data pipeline for a deployment of 10,000 sensors each emitting one 100-byte message every 10 seconds. What is the approximate sustained message rate the ingestion layer must handle?

  1. 10 messages per second
  2. 100 messages per second
  3. 10,000 messages per second
  4. 1,000 messages per second Correct answer.

Explanation

Each sensor emits 0.1 messages per second, so 10,000 sensors produce roughly 1,000 messages per second. Sizing ingestion capacity to this rate is a core planning calculation before adding headroom for bursts.

Question 6Planning IoT systems

Which device identity approach best supports scalable, certificate-based mutual authentication in a planned IoT fleet?

  1. Per-device X.509 certificates provisioned during manufacturing or onboarding Correct answer.
  2. A shared username and password embedded in all firmware images
  3. Disabling authentication to simplify initial rollout
  4. Relying on network-level MAC address filtering only

Explanation

Per-device X.509 certificates enable mutual authentication and allow revocation of a single compromised device without affecting the fleet. Shared credentials and MAC filtering provide weak or non-scalable identity assurance.

Question 7Planning IoT systems

In planning an IoT system, what distinguishes a gateway from a simple sensor node?

  1. A gateway must run the same firmware as the sensor nodes it serves
  2. A gateway always stores data permanently while sensors never buffer
  3. A gateway bridges local device networks to wider networks and can perform protocol translation and edge processing Correct answer.
  4. A gateway is only used for power distribution to sensors

Explanation

Gateways aggregate and translate between local device protocols and upstream networks, often adding edge processing. They are not defined by permanent storage or power distribution, and they typically run distinct software from constrained nodes.

Question 8Planning IoT systems

A planning team must ensure that sensor data remains trustworthy when transmitted over an untrusted public network. Which measure most directly addresses data integrity in transit?

  1. Using TLS or DTLS to provide authenticated encryption between endpoints Correct answer.
  2. Increasing the sampling frequency of the sensors
  3. Storing data in plaintext at the gateway before forwarding
  4. Compressing payloads with a general-purpose algorithm

Explanation

TLS/DTLS provides authenticated encryption that protects both confidentiality and integrity of data in transit. Compression, sampling rate, and plaintext storage do not by themselves guarantee integrity against tampering.

Question 9Planning IoT systems

When planning device provisioning at scale, which practice best reduces the risk of onboarding rogue devices?

  1. Using a trusted enrollment service with attestation or pre-shared enrollment credentials Correct answer.
  2. Allowing any device that connects to the network to auto-register without checks
  3. Relying solely on physical security of the deployment site
  4. Publishing the network join keys in public documentation

Explanation

A trusted enrollment service with attestation or pre-shared enrollment credentials verifies device legitimacy before granting network access. Open auto-registration and published keys invite rogue devices into the fleet.

Question 10Planning IoT systems

Which planning artifact most directly captures the expected behavior, inputs, and outputs of an IoT subsystem for later verification?

  1. A marketing brochure describing product benefits
  2. A list of preferred hardware vendors
  3. An interface specification or contract for the subsystem Correct answer.
  4. A network cabling diagram

Explanation

An interface specification defines expected inputs, outputs, and behavior, providing the basis for verification and integration testing. Cabling diagrams, marketing material, and vendor lists do not specify functional contracts.

Question 11Planning IoT systems

A deployment requires deterministic, bounded latency for safety-critical control loops. Which network planning approach is most appropriate?

  1. Randomized backoff on a shared unmanaged wireless channel
  2. Best-effort internet routing with retries
  3. A time-sensitive networking or industrial protocol with guaranteed latency bounds Correct answer.
  4. Batch processing of control messages every hour

Explanation

Deterministic control loops require networks with bounded latency such as TSN or industrial fieldbus protocols. Best-effort routing, randomized contention, and hourly batching cannot guarantee timing.

Question 12Planning IoT systems

In planning for data lifecycle management, why is defining a retention policy for IoT telemetry important?

  1. It controls storage cost and compliance exposure by limiting how long data is kept Correct answer.
  2. It eliminates the need for data encryption
  3. It guarantees that all data is deleted immediately after collection
  4. It removes the need for data backup strategies

Explanation

Retention policies bound storage growth and reduce regulatory and privacy exposure by defining how long telemetry is kept. They do not replace encryption or backup planning.

Question 13Planning IoT systems

A planner is comparing cellular IoT options for devices that need years of battery life and only occasional small uplinks. Which choice is most aligned with those constraints?

  1. Wi-Fi 6 with high-throughput video
  2. LTE Cat-M1 or NB-IoT Correct answer.
  3. Satellite broadband with continuous connection
  4. Standard LTE Cat-4 with continuous streaming

Explanation

LTE Cat-M1 and NB-IoT are low-power wide-area cellular options designed for infrequent small uplinks and long battery life. Cat-4 streaming, Wi-Fi 6 video, and continuous satellite links consume far more power and bandwidth.

Question 14Planning IoT systems

When planning a system that must operate in a remote area with intermittent connectivity, which design pattern best maintains data integrity?

  1. Reducing sensor accuracy to lower data volume
  2. Discarding data whenever the uplink drops
  3. Local buffering with store-and-forward and reconciliation on reconnection Correct answer.
  4. Sending data only during peak network hours

Explanation

Local buffering with store-and-forward preserves data during outages and reconciles it when connectivity returns, maintaining integrity. Discarding data or reducing accuracy sacrifices information rather than managing intermittency.

Question 15Planning IoT systems

Which consideration is most critical when planning the physical placement of IoT sensors in an industrial environment?

  1. Selecting sensor enclosures by color to match branding
  2. Placing all sensors in a single central cabinet for convenience
  3. Accounting for electromagnetic interference and environmental factors that affect readings Correct answer.
  4. Maximizing cable length regardless of signal degradation

Explanation

Industrial environments introduce EMI, temperature, vibration, and moisture that can corrupt readings, so placement must account for these factors. Aesthetics, arbitrary cable length, and centralizing all sensors can degrade measurement quality.

Question 16Planning IoT systems

A planning team needs to estimate the total cost of ownership for an IoT deployment. Which combination of factors belongs in that estimate?

  1. Only the purchase price of the sensors
  2. Only the cloud storage costs
  3. Only the labor cost of the initial installation
  4. Device hardware, connectivity subscriptions, cloud services, maintenance, and decommissioning Correct answer.

Explanation

Total cost of ownership spans hardware, connectivity, cloud/backend services, ongoing maintenance, and end-of-life decommissioning. Considering only one category understates the true cost.

Question 17Planning IoT systems

Which approach best supports scalability when planning an IoT backend that must ingest data from a growing number of devices?

  1. Horizontally scalable ingestion using partitioned queues or stream processing Correct answer.
  2. Storing all raw data indefinitely in a single relational table
  3. Manual periodic restarts of the ingestion service
  4. A single monolithic server sized for today's device count

Explanation

Horizontally scalable ingestion with partitioned queues or stream processing lets capacity grow with device count. A single monolith, manual restarts, and unbounded single-table storage do not scale reliably.

Question 18Planning IoT systems

A planner must decide between MQTT and HTTP for constrained devices that need efficient, event-driven telemetry with server push. Which choice is better justified?

  1. MQTT, because its publish/subscribe model and lightweight headers suit constrained, event-driven telemetry Correct answer.
  2. MQTT, because it eliminates the need for any security layer
  3. HTTP, because it supports server push natively in all deployments
  4. HTTP, because it always uses less bandwidth than MQTT

Explanation

MQTT's publish/subscribe model with lightweight fixed headers suits constrained devices and event-driven delivery, and brokers can push messages to subscribers. HTTP is request/response oriented and its headers are comparatively heavy for tiny telemetry.

Question 19Planning IoT systems

During planning, a risk assessment identifies that a compromised gateway could inject false commands into the control system. Which control most directly mitigates this risk?

  1. Increasing sensor sampling rates across the fleet
  2. Adding more gateways to increase redundancy
  3. Shortening the telemetry retention period
  4. Enforcing command authorization and signing at the control layer independent of the gateway Correct answer.

Explanation

Validating and authorizing commands at the control layer, with cryptographic signing, prevents a compromised gateway from injecting trusted commands. Redundancy and retention changes do not address command authenticity.

Question 20Planning IoT systems

A planning team must reconcile conflicting requirements: operations wants sub-second alerting, while finance caps bandwidth. Which design decision best satisfies both constraints?

  1. Removing alerting entirely to save bandwidth
  2. Reducing sensor sampling to once per hour
  3. Performing threshold detection at the edge and transmitting only alerts and summaries Correct answer.
  4. Streaming all raw high-frequency data continuously to the cloud

Explanation

Edge threshold detection delivers sub-second alerts locally while transmitting only alerts and summaries, minimizing bandwidth. Streaming all raw data violates the bandwidth cap, and reducing sampling or removing alerts sacrifices the operational requirement.

Question 21Planning IoT systems

When planning a phased rollout of an IoT system across multiple sites, which strategy best manages risk?

  1. Skip validation to accelerate delivery
  2. Pilot at a representative site, validate, then expand in stages with rollback plans Correct answer.
  3. Deploy to all sites simultaneously to shorten the schedule
  4. Deploy only to the least critical site and never expand

Explanation

A pilot at a representative site validates assumptions before staged expansion with rollback plans, containing the blast radius of failures. Simultaneous deployment maximizes risk, and skipping validation or never expanding defeats the rollout goal.

Question 22Planning IoT systems

A planner is evaluating whether to use a public cloud IoT platform or build a self-hosted backend. Which factor most directly affects data sovereignty compliance?

  1. The physical location and jurisdictional control of where data is stored and processed Correct answer.
  2. The number of sensors in the fleet
  3. The programming language used by the device firmware
  4. The color of the device enclosures

Explanation

Data sovereignty depends on where data is physically stored and processed and which jurisdiction governs it, which differs between public cloud regions and self-hosted infrastructure. Firmware language, fleet size, and enclosure color do not determine legal jurisdiction.

Question 23Planning IoT systems

Which planning practice best ensures that an IoT deployment can be maintained over a ten-year lifecycle despite hardware obsolescence?

  1. Avoiding documentation to reduce initial effort
  2. Designing abstraction layers and documenting interfaces to allow component substitution Correct answer.
  3. Locking in a single vendor with no exit plan
  4. Purchasing only the newest unproven components

Explanation

Abstraction layers and documented interfaces let planners substitute obsolete components without redesigning the whole system. Single-vendor lock-in, missing documentation, and unproven components increase long-term risk.

Question 24Planning IoT systems

A planner must define service-level objectives for an IoT monitoring system. Which SLO is most meaningful for detecting sensor outages?

  1. The average color temperature of dashboard displays
  2. The physical weight of each sensor node
  3. The percentage of expected sensor heartbeats received within a defined interval Correct answer.
  4. The number of firmware releases per quarter

Explanation

Heartbeat receipt within a defined interval directly measures whether sensors are online and reporting, making it a meaningful availability SLO. Display settings, release cadence, and device weight do not reflect sensor health.

Question 25Planning IoT systems

Which planning decision most directly reduces the attack surface of an IoT device fleet?

  1. Enabling every available network service by default
  2. Disabling unused ports and services and applying least-privilege access Correct answer.
  3. Broadcasting device identifiers openly on the network
  4. Using default vendor credentials on all devices

Explanation

Disabling unused ports and services and enforcing least privilege minimize the entry points an attacker can exploit. Enabling all services, default credentials, and open identifier broadcasting expand exposure.

Question 26Planning IoT systems

Which planning document defines the concrete, testable performance and availability targets that an IoT service must meet for its users?

  1. Firmware changelog
  2. Network cabling diagram
  3. Service-level objectives (SLOs) Correct answer.
  4. Bill of materials (BOM)

Explanation

SLOs are the planning artifact that states measurable reliability and performance targets, such as uptime percentage or alert latency, against which the deployed system is evaluated. They translate business expectations into numbers that operations can monitor and report.

Question 27Planning IoT systems

A planner is sizing a battery-powered field sensor that wakes briefly to transmit a small reading and then sleeps. Which design choice most directly extends its operational life?

  1. Polling the backend server every second for configuration updates
  2. Streaming raw sensor samples continuously over a persistent connection
  3. Increasing the radio transmit power to the maximum
  4. Keeping the radio and processor in deep sleep between scheduled transmissions Correct answer.

Explanation

For an intermittently reporting device, energy is dominated by time spent in active radio and compute states, so duty-cycling into deep sleep between transmissions is the primary lever for battery life. The design should minimize wake time and avoid unnecessary network activity.

Question 28Planning IoT systems

In an IoT architecture plan, which component typically terminates local device protocols and performs protocol translation before data reaches backend services?

  1. The certificate authority
  2. The cloud billing service
  3. The time-series database
  4. The edge gateway Correct answer.

Explanation

An edge gateway sits between constrained field devices and backend systems, terminating local protocols such as Zigbee or Modbus and translating them into IP-based protocols the backend consumes. This isolates protocol handling from core services and enables local filtering.

Question 29Planning IoT systems

Which planning practice ensures that an IoT solution can be reproduced identically across staging and production environments?

  1. Treating infrastructure and device configuration as version-controlled code Correct answer.
  2. Documenting configuration only in an email thread
  3. Manually configuring each server through an interactive console
  4. Allowing each site to choose its own untracked settings

Explanation

Infrastructure-as-code and version-controlled device configuration make environments reproducible, reviewable, and auditable, which reduces drift between staging and production. Manual or undocumented changes cannot be reliably repeated or rolled back.

Question 30Planning IoT systems

A planning team must choose how field devices will obtain initial network credentials without a technician configuring each unit by hand. Which approach best fits a zero-touch onboarding plan?

  1. Disabling authentication until the device reaches the field
  2. Shipping devices with a shared default password printed on the box
  3. Pre-loading unique per-device credentials during manufacturing and validating them at first connection Correct answer.
  4. Requiring each device to be manually keyed by an on-site engineer

Explanation

Zero-touch provisioning relies on unique credentials injected at manufacture and verified by a bootstrap service at first connection, so devices can authenticate without human intervention. Shared defaults or disabled authentication undermine the security of onboarding.

Question 31Planning IoT systems

When planning capacity for an IoT ingestion tier, which metric is the most direct indicator of whether the tier can keep pace with device traffic?

  1. Average size of the deployment's documentation
  2. Incoming messages per second relative to sustained processing throughput Correct answer.
  3. Total number of stored firmware images
  4. Count of registered user accounts

Explanation

Ingestion capacity is governed by the arrival rate of messages compared with the rate the tier can durably process, so planners size against peak messages per second versus sustained throughput. Other metrics do not describe message-handling load.

Question 32Planning IoT systems

A planner is defining how a fleet of sensors will be kept on accurate time so that readings from different devices can be correlated. Which planning choice best supports this requirement?

  1. Omitting timestamps and ordering records by arrival
  2. Timestamping records only when they reach the database
  3. Using a time synchronization protocol such as NTP or PTP with defined accuracy targets Correct answer.
  4. Relying on each device's unsynchronized local clock

Explanation

Correlating telemetry across devices requires a common time reference, so the plan should specify a synchronization protocol and the accuracy the use case needs. PTP can provide sub-millisecond accuracy where required, while NTP suits looser needs.

Question 33Planning IoT systems

During planning, a team must decide how to handle firmware updates for devices behind low-bandwidth links. Which strategy best balances update reliability with limited bandwidth?

  1. Disabling updates after initial deployment
  2. Pushing full images to every device simultaneously at peak hours
  3. Requiring each device to download from a different unverified source
  4. Using delta updates with staged rollouts and rollback capability Correct answer.

Explanation

Delta updates reduce the bytes transferred over constrained links, and staged rollouts with rollback limit the blast radius of a faulty release. Together they balance bandwidth efficiency with recovery from failure.

Question 34Planning IoT systems

A planner is selecting a data store for high-volume IoT telemetry that will be queried mainly by time range and device identifier. Which storage characteristic matters most in this decision?

  1. Ability to store large binary firmware blobs
  2. Support for full-text search of user chat logs
  3. Compatibility with desktop spreadsheet macros
  4. Support for efficient time-indexed writes and range queries Correct answer.

Explanation

Telemetry workloads are dominated by high-rate time-ordered writes and queries filtered by time window and device, so a time-series store optimized for those access patterns is the natural fit. The other characteristics address unrelated workloads.

Question 35Planning IoT systems

When planning redundancy for an IoT backend, what does an active-passive failover design primarily provide?

  1. Guaranteed zero data loss without any replication
  2. Elimination of the need for health monitoring
  3. A standby replica that can take over if the primary fails Correct answer.
  4. Higher write throughput than an active-active design

Explanation

Active-passive failover keeps a standby replica ready to assume service when the primary fails, improving availability at the cost of idle capacity. It does not by itself raise throughput or remove the need for health checks and replication.

Question 36Planning IoT systems

A planner must choose between a star and a mesh topology for field devices in a large industrial site with many obstructions. Which consideration most strongly favors a mesh plan?

  1. Mesh topology removes the need for any routing protocol
  2. Mesh devices can relay traffic around obstructions through neighboring nodes Correct answer.
  3. Star topology always consumes less power than mesh
  4. Mesh topology eliminates the need for security credentials

Explanation

In a mesh, nodes forward traffic for one another, so paths can route around obstructions that would block direct links to a single central hub. This improves coverage and resilience in cluttered environments at the cost of added routing complexity.

Question 37Planning IoT systems

A planning team is defining acceptance criteria for an IoT pilot before wider rollout. Which criterion best demonstrates that the pilot has validated the architecture?

  1. The pilot was approved verbally by one stakeholder
  2. The pilot used the most expensive hardware available
  3. The pilot met agreed latency, availability, and data-integrity targets under representative load Correct answer.
  4. The pilot ran for one day without any measurements

Explanation

Acceptance criteria should be measurable and tied to the SLOs, so demonstrating latency, availability, and data integrity under representative load shows the architecture can meet its goals. Vague or subjective approval does not validate the design.

Question 38Planning IoT systems

A planner is mapping data flows for a smart-building system and must decide where personally identifiable occupancy data may be processed. Which planning artifact most directly governs this decision?

  1. A hardware warranty schedule
  2. A data-flow diagram annotated with classification and processing locations Correct answer.
  3. A list of device serial numbers
  4. A marketing campaign calendar

Explanation

A data-flow diagram annotated with data classification and where each flow is processed shows which components touch personal data, enabling planners to enforce residency and handling rules. This makes it the artifact that governs processing-location decisions.

Question 39Planning IoT systems

A planning team must reconcile a requirement for local real-time control with a requirement that all telemetry be archived centrally. Which architecture best satisfies both?

  1. Send all telemetry to the cloud first and let the cloud issue control commands
  2. Buffer telemetry on devices indefinitely and never forward it
  3. Disable local control so that only the central system acts
  4. Perform control decisions at the edge and asynchronously forward telemetry to the cloud for archival Correct answer.

Explanation

Edge control meets the real-time requirement without depending on round-trip cloud latency, while asynchronous forwarding preserves central archival. This separates the latency-critical path from the data-retention path.

Question 40Planning IoT systems

A planner is assessing the risk that a single cloud region outage could halt an IoT service. Which planning measure most directly reduces this risk?

  1. Deploying the backend across multiple regions with tested failover Correct answer.
  2. Adding more dashboards for operators
  3. Increasing the number of devices per site
  4. Extending the telemetry retention period

Explanation

Regional redundancy with a tested failover procedure lets the service continue or recover when one region fails, directly addressing the single-region dependency. More devices, dashboards, or retention do not remove that dependency.

Question 41Designing IoT systems

A hospital wants bedside monitors to keep streaming vitals during a WAN outage and to buffer readings until connectivity returns. Which design element most directly satisfies this requirement?

  1. A round-robin DNS record for the ingestion endpoint
  2. Local edge processing with persistent store-and-forward buffering Correct answer.
  3. TLS mutual authentication between monitors and the gateway
  4. A larger cloud database instance with read replicas

Explanation

Resilience to WAN loss requires the data path to survive locally, so the design must include on-premises processing plus durable local buffering that forwards once the link recovers. Cloud-side scaling and DNS tricks do nothing while the uplink is down, and mutual TLS only secures a connection that no longer exists.

Question 42Designing IoT systems

Which IoT reference model layer is responsible for converting raw sensor readings into calibrated engineering units and filtering obvious outliers before transmission?

  1. The network transport layer
  2. The business integration layer
  3. The application and presentation layer
  4. The edge or device-level processing layer Correct answer.

Explanation

Calibration, unit conversion, and simple outlier rejection are local data-conditioning tasks that belong at the edge, close to the sensor, to reduce bandwidth and preserve fidelity. Transport layers move bytes, while application and business layers consume already-conditioned data.

Question 43Designing IoT systems

A design team must choose between a star topology and a mesh topology for 400 battery-powered environmental sensors spread across a warehouse. Which consideration most strongly favors mesh?

  1. Extended coverage and path redundancy when nodes can relay for one another Correct answer.
  2. Reduced need for any routing or neighbor discovery logic
  3. Guaranteed deterministic latency for every packet
  4. Lower per-node radio complexity and simpler firmware

Explanation

Mesh topologies let nodes relay traffic, extending range and providing alternate paths when a link or node fails, which suits large dispersed deployments. The trade-off is added routing complexity and energy cost, so the simpler star options describe advantages of star rather than reasons to pick mesh.

Question 44Designing IoT systems

Which protocol is the natural design choice when constrained devices must publish small telemetry messages to a broker over unreliable links with minimal header overhead?

  1. HTTP/1.1 with keep-alive
  2. MQTT Correct answer.
  3. SNMPv3
  4. FTP

Explanation

MQTT uses a compact publish/subscribe model with small fixed headers and quality-of-service levels designed for constrained, intermittent links. HTTP carries heavy per-message overhead, while SNMP and FTP are management and file-transfer protocols unsuited to high-volume telemetry.

Question 45Designing IoT systems

What is the primary purpose of assigning each device a unique, immutable identity credential such as a per-device X.509 certificate during manufacturing?

  1. Enabling strong authentication and revocation of individual devices Correct answer.
  2. Compressing telemetry payloads before transmission
  3. Reducing the power consumed by the radio
  4. Increasing the wireless range of the device

Explanation

Per-device credentials allow the platform to authenticate each device individually and revoke a single compromised unit without affecting others. They have no effect on radio range, power draw, or payload compression, which are physical and application-layer concerns.

Question 46Designing IoT systems

An engineer needs devices on a factory floor to receive firmware updates in a bandwidth-efficient way when many identical units are behind a single gateway. Which design pattern fits best?

  1. Multicast or group-based distribution from the gateway to devices Correct answer.
  2. Emailing firmware images to operators for manual installation
  3. Streaming the image continuously in a loop to all devices
  4. Each device independently downloading from the vendor cloud

Explanation

Group-based or multicast distribution lets the gateway fetch an image once and fan it out to many identical devices, conserving WAN bandwidth. Independent cloud downloads multiply WAN traffic, and manual or looping approaches are inefficient or unsafe.

Question 47Designing IoT systems

Which practice best supports data minimization in an IoT design that collects occupancy data in an office building?

  1. Storing raw data unencrypted to simplify analytics access
  2. Retaining all raw sensor streams indefinitely for future use
  3. Aggregating counts at the edge and discarding raw identifiable traces Correct answer.
  4. Sharing raw streams with all third-party vendors by default

Explanation

Data minimization means collecting and keeping only what the purpose requires, so aggregating at the edge and discarding raw identifiable traces limits exposure. Indefinite retention, unencrypted storage, and broad sharing all expand risk and contradict the principle.

Question 48Designing IoT systems

Why is a device shadow (digital twin of device state) useful in a design where devices sleep most of the time?

  1. It lets applications read and write last-known state without waiting for the device to wake Correct answer.
  2. It guarantees sub-millisecond actuation latency
  3. It permanently replaces the need for the physical device
  4. It encrypts all traffic between the device and the cloud

Explanation

A device shadow stores the last reported and desired state in the cloud, so applications interact with the shadow while the sleepy device is offline and the device syncs on wake. It does not replace hardware, encrypt traffic, or guarantee real-time actuation.

Question 49Designing IoT systems

A fleet of vibration sensors samples at 20 kHz, but only anomaly features are needed upstream. Which design decision most reduces backhaul cost while preserving detection quality?

  1. Batch raw samples hourly into large uncompressed archives
  2. Increase sampling rate to 40 kHz and forward everything
  3. Compute spectral features on the edge and transmit only feature vectors Correct answer.
  4. Transmit raw waveforms continuously to the cloud

Explanation

Extracting spectral features at the edge collapses high-rate raw data into compact vectors that retain the diagnostic information needed for anomaly detection. Sending raw or higher-rate data, compressed or not, multiplies bandwidth without adding value to the upstream model.

Question 50Designing IoT systems

An architect must ensure that a command to close a valve is acted on even if the first delivery is lost, while avoiding duplicate actuation. Which combination best satisfies both goals?

  1. Best-effort fire-and-forget publishing
  2. At-least-once delivery with an idempotent command handler keyed by message ID Correct answer.
  3. Broadcast to all devices without acknowledgement
  4. At-most-once delivery with no message identifiers

Explanation

At-least-once delivery with retries ensures the command eventually arrives, and idempotency keyed on a message identifier prevents a retried command from actuating twice. At-most-once can lose the command, and fire-and-forget or broadcast offer no delivery guarantee.