Font size
WorksheetsIntroduction to AIoT Topics
Total questions: 30
Worksheet time: 15mins
A hospital wants to reduce emergency response times by integrating AI with its IoT sensor network. Which approach best leverages AIoT principles to achieve this goal?
Stream sensor data to a dashboard for manual triage decisions
Use AI models on edge devices to detect anomalies and trigger alerts
Replace IoT sensors with higher‑resolution cameras for richer data
Batch upload daily logs to a cloud server for retrospective analysis
A hospital wants to reduce surgical complications and tailor treatments to individual patients. Which AI-driven approach best addresses both goals while using evidence from patient data?
Personalized medicine with robotic surgery support
Standardized treatment plans for each disease
Manual record reviews by senior physicians
General symptom checklists for all patients
A bank observes rising chargebacks in online payments. Design a strategy using AI that reduces losses while maintaining customer experience. Choose the most defensible plan.
Use fraud detection with risk scoring and chatbots
Delay all transfers for a full business day
Increase manual audits of all customer accounts
Disable card-not-present transactions at peak hours
A logistics firm struggles with delivery delays in a congested city. Which AI application would most effectively improve on-time rates through planning and continuous feedback?
Predictive maintenance for warehouse robots
Inventory optimization for retail partners
Image analysis for package labeling
Autonomous vehicles with traffic management
A school district plans to reduce grading time while still improving feedback quality for diverse learners. Which combined AI approach best meets this goal?
Personalized tutoring without assessment data
Chatbots answering generic curriculum questions
Adaptive assessments with automated grading
Automated grading with fixed rubrics only
A city plans to retrofit old water meters so they can report leaks as they happen without manual checks. Which upgrade most directly aligns with IoT principles to achieve this goal?
Use manual logs sent by email once per month
Store readings locally without any network connectivity
Schedule weekly technician visits for inspections
Attach sensors and network modules for real-time data
Fill in the blank: The main goal of IoT is to transform a device without processing capabilities into a (a) device that can exchange real-time data over the internet without human involvement.
A factory wants to reduce unplanned downtime using sensors on machines. Which IIoT-enabled approach best fits this goal, and why?
Increase production speed to reveal weak components
Use predictive maintenance to anticipate failures
Run periodic manual inspections to detect faults
Replace all machines at fixed calendar intervals
An engineer must justify investing in IIoT for a power plant. Which reasoning most accurately explains IIoT’s focus and capabilities in this context?
IIoT is general consumer IoT for home devices
IIoT connects smart industrial assets to analyze data
IIoT is a cloud-only service for office automation
IIoT replaces all human operators with robots
Quality issues appear randomly on a manufacturing line. Choose the IIoT-driven plan that uses evidence from connected devices to improve outcomes.
Add more workers without data collection
Use fixed schedules for maintenance only
Shut down the line every few hours
Enable HMI analytics for quality control
A factory is migrating from small wearable sensors to mission‑critical robotics and power systems. Which shift best explains why the network must adopt more sensitive sensors and stricter controls?
Wearables consume less battery power
Power systems require cloud-only storage
Risk becomes life‑threatening on failure
Robots need faster Wi‑Fi standards
An enterprise architect must choose between IIoT and consumer IoT for a city’s railway and logistics network. Which option aligns with the required network scale and product life cycle?
IIoT with large scale and long life
IoT with small scale and long life
IoT with large scale and short life
IIoT with small scale and short life
You are planning data pipelines for two deployments: home wearables and industrial machines. Which pairing of data volume and security is most appropriate?
IoT: very high data, identity privacy; IIoT: medium‑high data, robust security
IoT: medium data, robust encryption; IIoT: very high data, identity privacy
IoT: low data, no security; IIoT: medium data, basic passwords
IoT: very high data, no privacy; IIoT: low data, strong privacy
A smart factory suffers frequent delays because sensor data must travel to a distant cloud before analysis. Which redesign best addresses this latency while enabling autonomous responses on-site?
Increase bandwidth to the central cloud
Batch data hourly for cheaper processing
Move ML inference to edge devices
Retain cloud analytics, add redundant sensors
Plan an AIoT solution to turn raw IoT data into actionable insights for home energy management. Which sequence is most effective and feasible?
Collect locally, apply on-device ML, trigger automation
Aggregate centrally, manual review, adjust settings weekly
Encrypt data only, store on gateway, no processing
Stream to cloud, archive logs, email monthly reports
You are evaluating an IoT deployment that struggles with massive data volumes, limited scalability, and no autonomous decision-making. Which prioritized roadmap best mitigates these constraints?
Replace sensors, postpone analytics indefinitely
Deploy edge AI for decisions, scale distributed processing
Migrate all components to a single cloud region
Add privacy policies, keep current architecture
A smart factory wants near-instant anomaly detection on vibration data, while still performing weekly model retraining and long-term storage. Choose the best placement of tasks across edge, fog, and cloud to meet latency and scalability needs.
Inference at fog, training at edge, orchestration in cloud
Inference in cloud, buffering at fog, storage at edge
Training in cloud, inference in cloud, storage in fog
Inference at edge, preprocessing at fog, retraining in cloud
A city’s traffic system must respond to incidents in under 100 ms while networks are occasionally congested. Select the strategy that best balances communication reliability and decision latency.
Batch uploads to fog every 10 minutes
Streaming all raw video to cloud for analysis
Centralized reinforcement learning with periodic commands
Local sensor fusion and actuation at edge nodes
A factory wants predictive maintenance that labels past failures but faces unlabeled streaming sensor data. Which learning strategy best combines these to improve models over time?
Unsupervised clustering on live sensor streams
Pure supervised training on historical failures
Rule-based thresholds without machine learning
Semi-supervised learning leveraging few labels
You are tasked with detecting abnormal energy usage in a smart building with no labeled anomalies. Choose the most appropriate approach and justify the choice based on algorithm properties.
Deploy k-NN using known fault labels
Apply k-Means or DBSCAN to find outliers
Train logistic regression with labeled anomaly data
Use decision trees for labeled classification
An IoT health monitoring system must both forecast failures and self-optimize. Which paired capabilities from AI in IoT directly support this requirement?
Anomaly Detection and Personalization
Predictive Analytics and Autonomous Control
Context Awareness and Personalization
Autonomous Control and Anomaly Detection
An industrial IoT team has very few labeled fault samples but streams large volumes of unlabeled sensor data. Which learning strategy best balances data constraints while improving fault diagnosis performance?
Supervised learning with extensive labeled datasets
Reinforcement learning through reward-driven exploration
Unsupervised learning without any labeled guidance
Semi-supervised learning combining labeled and unlabeled data
You are designing a camera-based traffic monitoring device. Choose the most suitable deep learning approach and justify the choice by the data modality described.
Autoencoders for tabular demand forecasting data
Policy gradient methods for reward maximization
Convolutional neural networks for image and video analytics
Recurrent neural networks for sequential text inputs
A smart building controller must learn optimal HVAC actions through trial and error within the building. Identify the key RL components the designer must formalize to enable this learning.
Labeled examples and class boundaries
Hidden layers and activation functions
Agent, environment, states, actions, rewards
Word embeddings and sentiment lexicons
Design an AIoT customer support kiosk that respects privacy while giving instant voice responses. Choose the most suitable approach.
Use rule-only expert system without speech module
On-device speech recognition with local intent parsing
Record audio and analyze later in remote servers
Text transcription in cloud with periodic replies
In a battery-powered industrial sensor, the inference component should be (a) to meet power limits while keeping latency low.
An engineer needs robust diagnostics in a cyber-physical system by combining learned patterns with explicit rules. Which technique aligns with this plan?
Hybrid AI models merging ML, DL, and symbolic AI
Standalone deep neural network trained end-to-end
Pure rule-based expert system without learned models
Manual threshold tuning by domain operators
A city plans an AIoT rollout to improve traffic flow and energy savings without constant cloud dependence. Which architecture best meets these goals while balancing power limits and security risks shown in the image?
Centralized cloud analytics with periodic batch uploads
Edge processing for local decisions with secure gateways
On‑device logging only with monthly manual transfers
Hybrid cloud streaming for all data with VPN tunnels
A warehouse robotics team must reduce downtime and maintenance costs while ensuring interoperability across mixed vendors. Which integrated AIoT strategy is most appropriate given the applications and challenges depicted?
Isolated robot controllers with no shared telemetry
Single vendor lock‑in to avoid protocol translation
Manual inspections scheduled by calendar reminders
Predictive maintenance models on IoT sensor streams
An engineer observes noisy sensor data in an automated vehicle. To maintain safe real‑time decisions under power constraints, what should be prioritized first?
Frequent model retraining during driving sessions
Disabling preprocessing to reduce latency
Lightweight filtering and validation at the edge
Higher‑resolution cameras streamed to the cloud
